Moche Valley, Dong with Spondvlus Shelis

THE QUEST FOR MULLU: CONCEPTS, TRADE, AND THE ARCHAEOLOGICAL DISTRIBUTION, OF SPONDEUS IN THE ANDES

A Thesis Subrnitted to the Cornmittee on Graduate Studies in Partial fulfilhnent of the Requirernents for the Degree of Master of Arts in the Facuiîy of Arts and Science

TRENT UNMZRSITY Peterborough, Ontario, Canada

CI Copyright by David Blower 1995

Anthropology MA. Program May 19% The author bas grimtbd a non- L'auteur a accon% me licence non exclusive licence ailowing the exc1uSivepum~Bla NaticmaiLi'braryofCrmedato Biblioth&penationale du Canada de reproduce, 1- distnUorsell Iepmduk,pr&r, ~'buerou copies ofhïskr fbesis ôy any means vedre des copies & sa thbe & andhanyformorfamiat,makhg this thesis available to interested fomque ce soit pour mettre des pefsO= exemplaires de cette th&= à la

The author retains omdpof the L'auteur comme la proprieté du copyright m hisha thesis. Neither droit d'auteur qui protège sa thèse. Ni the thesis nor mbstanM exîmds &om it may be prÉated or otherwise reproduced with the author's permission. ABSTRACT The Quest For Mullu: Concep- Trader And The Archaeologïcai Disnibution, Of Spondvlus In The Andes

David Blower

This thesis is desiped to ident- the nature of the relationship between Spon&lus. a sea mollusk and mullu a Precolumbian Andean concept, with the intention of undemandino the sheli's value and meanM in a system of made and exchange. By dennino the riaial and ceremonid value of the shell it is possible to undentand why it was an important trade commodity, and why it might have been the foundation on which a trade network was based. Through a review of the available literature and original sources, and an anaiysis of the shelPs appearance and context in archaeological sites' the concept of mullu is defined as more complex than previously considered. It is seen to hction on different levels of ritual ceremony and in varied contexts and associations. The importance ofmuilu to trade networks and the ideology of the Andes foms the basis for its designation as a rinial and trade complex. This thesis would not have evolved into its present state without the support and contributions of supervisory committtee who read and commenteci on the eariier drafts: Dr. John Tepic‘ Dr. Theresa Tepic‘ and Dr. Joan Vastokas. Dr. John Topic has supe~sedthe deveiopment of this work corn its early dr& to completion, but both he and Dr-Theresa Topic have been involved completely in my exposure to Andean archaeology, through course work and field study, since my time as an under-mduate. The two field seasons of work in Ecuador which led to this study were conducted under the support of gants awarded to both Dr. John Topic and Dr. Theresa Topic. To both of them 1 express my appreciation for the opportunity to work in the Andes and the chance to observe and experience the culture firsthand. Thanks also to Dr. Daniel Sandweiss who not on- acted as my extemal examiner, but who was able to recornmend source material which afEected my understanding of pre-Columbian open-sea navigation and the issue of Chincha trade. ïhe information provided through these additionai references was greatiy apprec iated. During- the two seasons spent in Ecuador 1 was fomuiate enough to meet two persons who could introduce me to local customs, discuss their own experiences in Andean archaeology, and who were very patient with my attempts at Spanish conversation. in the nrst season, ALfredo Melly Cava fkom Trujillo, Pen: and in the second Jhonny Villon Mordes, Guayaquil, Ecuador, provided companionship and fnendship during my stay. In both instances theu efforts helped -preatiy with my pneral understanding of Latin Arnerican culture?and also opened a few doors that would have remaineci closed to a novice gringo. While examinino the collections of Spondvlus material at various Ecuadorian institutionst 1 was to meet many individuds who offered heip and encouragement I would like to th& two in panicular. In Quito. Rosa Ordofiez Romero, Auxiliar 'ïécnico, at the Museo Sijon Y Caamailo. Universidad de Catolim was very helpfid with my examination of the Spondvlur figurines and artifacts hmCerro Namio. During my coastal travels to visit archaeoiogical sites and collections related to the early use of Spndvlus 1 had the pleasure of meeting Sr. Olaf Holm, former Director of the Museo of the Banco Central in Guayaquil. Sr. Hohoffered encoqement and information on some of the more obscure references to muilu and Spondvius. for which 1 mi@ still be searching, and who has continued his support through correspondence since my departure kmEcuador. The contributions of the people mentioned? and others not mentioned. have had an effect on the final appearance of this thesis, but without the support of my wife Vem and our daughters Devon and Morgaq 1 am doubtful that there would be a thesis at ail. To them goes my deepest appreciation and love. While the above individuds have greatly &êcted the final thesis, they are in no way responsible for any omissions, misconceptions. misunderstanding or other errors that appear in this text 1 have always been capable of making those mistakes on my own. TABLE OF CONTENTS

Chapter Page

ONE: Introduction

WO Spondylus Spondvlus And Its Habitat Spon4viw princeps Spondvlus princeps princeps Spondvius princeps leucacantha Spon&ius princeps cdcifer Spondvlus unicoior Spondvlus Physiology Deteminino the Clrigin and of Spondylus Material

TmEEi Mullu Onho-nrap hic Variation Things That Are Mullu Technicolour Muliu: Colour in The Chronicles Gender?Duality And The Vagina Dentata Rab Lakes, And Springs: Water Association Food For The Gods. Food For The People Mullu Swival In The Colonial Period And Beyond s-ary FOUR Procuremen& Transportaào~and Tmde Spondvhîs Procurement Transprtation: Delivering The SheII ïhe Vaiue of Spondvlus The Mulïu Trade: Building A Network Maritime or Land Trade: Amcuiating Systems Chincha: The Centre of Trade? The Mesoamerican Comection System Operators: Trade Specialists Summq

FISE Spondylus Distribution Archaeolo@cal Distributions and the Evidence of Trade Ecuador: From Coast to Hi-@lands Petq Over Theand Space Summq

SIX Conclusions

References

Figures

Appendices A Andean Sites Containing Spondvlus Materials

B Toponyms Containing Variations Of The Word Mullu In Ecuador

C SpondyZus Material In The Museo de Jacinto Jijon y Caamaiio (Quito)

D The Use of Spo~lusIn Mesoarnerica vii

List 0fFog.res

El Segundo MW Febrero. Pavcar Varai QuilIa 139 idolos 1 Vacas De Los Chinchai S~YS.Los yndios 140 idoles I Vacas De Los Andi Svios. SacrEcauan Ande Suyo 141 1dolos 1Vacas De Los Colla Svios 142 fdolos 1 Vacas De Los Conde Svios 143 Paccha in the fonn of a plate of Spondjlus hmChimu 144 Spondvlus princeps 145 Spondvlus princeps leucacantha 146 S'ondvlus caicifer 146 Spondylus princeps unicolor 147 Spine pattern differeaces on princeps unicoior and prntcepsprinceps 148 Roll-out of the TeHo Obelisk 149 The Smiling God fiom Chavin de Huanthr 150 Staffgod (C-1) 151 Staffgod (C-4 ) 152 Coricancha cornogram by Santacruz Pachacuti Yamqui 153 Staff Goddess i~iustra~gthe vagina dentata 154 SWGoddess showing variation of the vagina dentuta 155 Spn@Ius sheil diving scene fiom a Middle Sican silver earspool 156 Middle Sican gold earspool with Spondyius diving me 157

Middle Sican -Id earspool Spon~vlusshell diving scene - 158 Middle Sican metal earspool SpondvIus divin$ scene 158 Middle Sican silver cylinder showïn~Spondjdus diving 159 Diving motif hmthe North Coast of Peru Middle Sican spatuia showa Spondvlus shell diver MtlZZu Toponyms in Ecuador Hatvn ChasqM. Chvrv Mvllv Chesqvi .Cvraca. Chm rnullo chasqui Eciirrtinrianbalsa& Construction view of sea-going vessei Dimibution of Spn&&tls habitat and routes ofSpm&I~exchange Andean Spondylus Sites List of Mole and Worked Shell by Site Figurines fkom Cern Narrio Ecuador Chronolow Peru Coastai Chronology Pem Himand Chronology CHAPTER ONE

INTRODUCTION

The cosmology of the Precolumbian Andes incofporated many beliefs and materials of ritual or sacnficial importance. One such materiai was Spondylus? a wann water sea moilusk whose naturai habitat extends fiom the GuIf of California to Ecuador. Important to Ptecolumbian Andeans for its use in dedicatory offerinos and sacrifice, Spondylus is found in archaeolo@cal sites thtoughout the coasts and highlands of Ecuador, Pew Bolivia Chile and into Argentina. However, it is not unique to this area. 'Ruoughout the Americas Spondvlus appears in archaeological contexts that illustrate its valued position and function across cultural boundaries. In Mesoamerica Spondylus appears in situations sirnilar to that of the Andes* but its presence in the prehistoric cultures of North Americ* Melanesia and Europe give evidence of a more universal relationship bebveen people and shells fkom the sea (Claassen and Sigmann 1993; Haistead 1993; Shackleson and Etderfield 1990). In the absence of written records, the role that this shell played in Andean cosmology is pieced together in this thesis through a compilation of ethnohistoric chronicles, indigenous oral accounts and the published results of archaeoiogical investigation. While Spondvlus shell is a recoverable artifact that can illustrate patterns of use over space and tirne, it is the Quechua tem mullu, as it refers to Spandvius. that mua be defhed and understood. Both the physical and conceptuai definitions of Sporuiylus are linked to each other, and are important to an understanding of the mullu concept, but at tirnes the word mullu and the physical object referred to as SpondvIus may differ- The question "if Spondylus is mullu is mullu necessarily Spondvltst' must be answered. Are the two terms as synonymous as they first appear to be, or are there conceptuai associations to rnullu that make it more complex than just an equivdent linguistic gioss for Spondvlzîs? 1 wÏIl argue here that mullu describes a concept broder than Spon4vizîs and that use of the two tems as synonpous is inappropriate and confushg. This confusion began with the initial interpretation of nullu by the chroniclers as a sea shell used by the natives to make red and white beads (Sihano 1%7:66 [1525]). Their interpretation of the tenn failed to recogize the complex of concepts behind the word, treathg the shell itself as rnullu instead of the function and meaning which it represented. It is understandable that this confusion occurred, since while the chroniclers were leaming the Quechua vocabulary of the Andes: both the use and signifieance of Spondylus were declining. Almost five hundred years later we are still wing to define the concept of mullu. Mullu as Spondylus held a rnultifaceted position in the ideology of the Andes. A symbol of sexuality, agricultural and human fertility, and rain, it was offered as a sacrifice to the sprlng and rivers that brought water (Davidson 198 1330; Marcos 1986: 197; Murra 1975). For some ceremonies it was sacrificed in conjunction with children. animais and other items such as gold, silve- chicha coca and fders (Guaman Poma 1980:213,221,225z 239; see Figs. 1-51 Molina 1989 115751: 121; Munia 1987 [1590]:420). At thes, the sacrifice of mullu was made as an offering for good health (Molina 1989 [1575]: 133). On some occasions the ground up shell and chaquira were thrown into fiurows to promote agricultud fertility, and at other times crushed Spondvlus shell was scaaered on the ground where the lord would walk (Cabello Valboa 195 1 [1586]:327). It was mixed with corn flour and the blood of sacrificed animais added to chicha as a drink offerhg1 (Acosta I962:248; Avila 1975 [1598]:5L; Carrion Cachot t955:38; Munia 1987 [1590]:422), and it was used whoie as dedicatory offerings in the post-holes and tloors of ceremonid buildings. As a concept, dluwas representative of the Andean system of duaiity and complementary oppositions such as upperllower?leWright and male/female. In one early Spanish transcription of an Andean myth muïZu is recopized not oniy as an offering to the go& but also as the food of the gods2 (Avila 1975 [1598]:52). These are just some of the uses and meanings associated with mulluF a concept which wiil be defined Merthrough associations described in the chronicies and native imaoe~and supported by the evidence of the archaeological record. In archaeological conte- Spondylus appears in both worked and unworked forms, and is found as burial accompaniments or dedicatory offerines. Its appearance in burials as pendants or necklaces may indicate additional use of the shell as a status marker often associated with female skeletons- The use and importance of mullu in Andean ritual are evidenced by the visual representations of the Spondvlus shell itself. From diving and the harvesting of Spondvlus to its representation in design motifs, Spondvlus shell appears as a recurring image in artwork ranging £tom Chavui (Burser and Salazar- Burner- 1993): Carhua (Wallace 199 la) and Middle Sich (Cordy-Collins 1990) to the Chimu at Chan Chan (Davidson 1980; Pillsbw 1993).

1 "...Usande estas conchas cuasi en todas las maneras de sacrificios, y aun el dia de hoy echan algunos el mollo molido en la chicha por supersticion. (Acosta 1%2248).

2 "...En ese momento, vino un hombre ilorando intensamente; traia a su hijo; llcvaba tambicn mulla, coca y un potajc sclccto liamado ticti, "Para quc los tomc HualiaiIo [Ca~incho]",decia" (Avila 1975 [1598]:51) ïhe visual record is part of the context on which the concept of mullu is based. The archaeological artihcts adomed with Spondvius material and representations of Spondvlus shells tell us that this sheII was important at specific time periods and locations diroughout the course of Andean prehistory During the Late Intermediate Period the door to the house of the id01 at Pachacamac was decorated with ground dom Spondjdus princeps shells (Pérez Bo- 1975321). Cupisnique stimip putvessels and Chhu/Inca plates fiom the North Coast (Camion Cachot 195538 Davidson 1980; Pérez Bonay 1975: I4IJ78) ail show variations of the Spondylus shell theme (Fig. 6). As important as fimding Spon&lus in the archaeological record is for locating possible trade systems? ritual activity and methods or places of production. the visual representations indicate its place in the belief -stem and its status as a valued object. For the purpose of this thesis. the term mullu is used to express the Quechua meaning of Spondylus as a concept that combines its physical and ideological attributes. A singie definition of these ternis is not possible as the meanin~c. may change when used in difEerent geopphic regions, different ritual contexts, and in conjunction with other words. An important side-efféct of Spondvlus research concems its value as an indicator of trade or exchange. The ideological implications of Spondvlus should not overshadow the economic information available conceming the methods of its procurement and transportation. How it was used, and its inherent value due to those uses and needs? may have been responsible for the formation of long distance trade networks. The by-product of this use is the information that allows us to rebuild these possible networks. This thesis is divided into three sections: the fim concems the Spondvlus shell itself. its physical characteristics and atmbutes the second examines the mulïu as a concept and the evidence through which we identm its fimction and meaning and the third considers the evidence of its procurement, transportation and made. Throughout these three sections' however, those aspects of Spondvlus use that indicate ritual and trade activity, and its conceptual identification as muilut wil1 be examined Whether or not Spondylus and nullu are synonymous terms which refer to shell materiai and chrrquira beads, or whether mullu incorporates a larger? symbolic/rnetaphoric category of meanings and activities that cm be identifiecl as a rÏtuai and trade cornplex will be discussed in the following chapters. CHAPTER TWO

Description of the physicai characteristics of Spondvlus shells and species is important to the identification of different cultural uses and the detemination of several possible levels of cultural rneaning. AI1 Sponàylus sheils do not appear interchangeab- in the archaeological record. Whether or not one species was preferred over another for specific functions or utilïzed subject to availability is unclear. Idenwing the shell as to its species and sub- species provides information on two distinct aspects of this study: first: it indicates the preference of use and the different values associated with each species, and second, it enables researchers to narrow down the geographic area from which the shell originated for use in the identification of trade.

Spondyius and Its Habitat

Spondvlus is a genus of mollusk in the Spondylidae family that is distributed worldwîde in tropical waters from the Atlantic Coast of the Amencas and West Indies to the Mediterranean, Hawaii to the Philippineso Australia and Japan, and the Indian Ocean. It is this brosd availability of Spondvlus to human populations that partial- accounts for its presence in the archaeological sites of Australi& Europe, North Amerka and South America (Claassen and Si-mann 1993; Halstead 1993; Shackleton and Elderf?eld 1990). The context of its presence in these sites suggests that the human populations of dl these areas associated Spon&ïus with simiiar patterns of socioeconomic and rituai use to that of the Andes. In the Rhine and Danube Basins of Neolithic Europe Spon&lus a~pearedin the forms ofnecklacesrpendants and bracelets (Clark 1986:8) more than 1700 km fiom its source in the Aegean Sea As in the Andes. Empan Spondvlus indicates some form of exchange network(s) due to the distance travelled. Only two species, Spondvlus princeps and Spondvlus calcifer. wwhh are found in the Pacifïc coastai area known as the Panamic Province are of interest to this study of Andean Spon&lus. The Panamic Province? more precisely known as the Panamic-Pacific Province, covers the zone of shallow-water fauna that extends appro'umately 7250 kilometra hmthe Gulfof California at latitude 30' 30' North, dom to latitude 4' 15' South at Cabo Blanco in northwestem Peru Olsson 1% 1:24). The differences between these two species are based on morphological attributes, and in the case of sub-species, on their range and naturai habitat. Although both species are referred to interchangeably by the general term Spondvlus. the spiny projections that are present on princeps have caused it aione to be known as the thorny oyster: or ostra espinosa (Norton1986:133). Whether calcifer and princeps are used interchangeab- and accorded equal value in ritual or monetary contexts is a question to be discussed in this thesis. The species and sub-species as described by Keen (1971 :96-98). Lamprell (1987:56.60), andolsson (1% 1: 150-153). are listed below. Spondvks princeps is a bivalve heid together by a ligament and an interlockhg hinge. The outside morphology of the shell, known as the sculptureo is fomed by radiai rows of crowded spike-Iike spines and the primary interspaces that are covered by secondary and temary spines. ïhe average size can reach 130 mm - 150 mm, but occasionaily lqerexampies appear. The colour of the outer shell varies fiom orange to coral-red. Variation in the positioning of secondary and tertiary spines and the colour of the valves and spines is illustrated in the three sub-species that follow.

Spndyksprinceps princeps (Figure 7)

This is the most common of the princeps sub-species to appear in the Andean area due in part to a habitat that ranges from the Coast of Panama in Central America south to northwestem Peru. The valves are usually regular in shape with mqteeth dong the outer margins. The outer shell has crowded, close-set short or medium-len-gh, spike-like spines projecting away from the shell. The primaiy and secondary spines are set in six rows. Additional temary spines appear in pairs in the space between the primary and central secondary rows. The primarv' secondary, and tertiary spines can be identified by their diminishing size. Each interspace between the primary rows has one central row of secondary and usually four rows of temary spines. two on each side of the middle. It is this arrangement of spines and the amount of space between rows that is important to the identification of sub-species. On the inside of the shell there is a wide and deepiy coloured marginal band that pdlels the outer edge of each valve. The extemai coiour of the sheil body and spines can be orange or coral red. At tintes this shell is referred to as Spo~vZuspictorwn,a name that appears in early archaeological literatwe and is stili in use in some modem South Amencan museums-

SpondjÇus pn'mceps Icucaœntha (Figure 8)

Possibly the oniy sub-species of princeps that is native to the Coast of Ecuado- leucucantho may only appear to be a dif5erent species due to its adaptation to difEierent environmental surroundïngs. Physicaliy, the spines on the outer shell are less crowded and the interspaces more open than other princeps varieties. The primary spines are longer than on princeps princeps and fluted or foliated at the ends. The inner margin is deeply coloured like the other varieties. but the spines are white or yellow and Iighter in colour than the main body of the sheil. Unfommately, the meends of the spikes are ohmissing in archaeological contexts due to either damage in transit' or the requirements of the finished product. ïhis makes identification of sub-species much more difficult.

Spondylùs calcifer (Figure 9)

Ranging fiom the Gulf of California south to Pem this species has many similarities to princeps but is easily identifiable by the purplish colour of its outer shell when present on artifacts. The main difference is ske, as it can grow to an average 155 mm or more across and weigh up to 1.5 kilogams. Spondylus calcifer displays srnail spiny projections in its early stages that are more numerous and evenly distributed than on princeps. The mature calcijèr specimens do not have spines, but the presence of spines in the immature s~ecimensleads to confusion when differentiating between the two species. Apart from the lack of spines and larger skcalcifer ahdiffers in the shape and sue of the two vaives. The vaives are asymmemcai with a border of foliated concenmcs, growth rings that can be seen as layes on the outer edge of the shell. The inner shell is white with a purplish band on the inside mar-- and an outside colour of purple or violet. This species is refened to as Sponàyhs ïimbatus in early literature and some museum collections.

Spondylus princeps unicoior Figure 10)

The inclusion of a sub-species that is normally found corn the Gulf of California south to Jalisco, Mexico. is important partly because the rmse of these shells has not been defined with any degree of certaintv. and partiy because its physicd differences allow for species cornparison relating to point of origin. When it is final- possible to identiQ Spondvlus matenal as to its point of ori-@n? the ability to recognize West Mexican Spondvlus in Andean sites will contribute important information to the question of whether or not Ecuadorian traders au-pmnted dieu supplies with irnported foreign shells. The unicolor valves are rounded and synmetrical in shape with primary spines that are enlarged but often flattened or fluted between wide, open interspaces. There are three smaller rom of spines between the prïmary rows -- one secondary and two tertiary on either side of the second-. The colour is unifonn for spines and shell body with an interior that is porceilaneous white and a deeply coloured band. The basic difference between this shell and princeps princeps is the number of te- spines. Unicolor has one on each side of the secondary spine and princeps has two (Figure 1 1). identification of Spondvlus is difficult? as no two specimens of any species are exactiy alike and can differ in shape, spine development, and colouration. depending on the Limitations of space, nutrition and water rnovement (L,amprell1987:9). As a bivalve, the Spon@Ilus mollusk has a bal1 and socket type hinge with a centrai ligament and two large interlockhg teeth on each side. The valves are held together by a locking mechanism that prevents the shell fkom being opened by predators. This mechanism includes the adductor muscle, denticulations on the edges of the two valves that interlock and the hinge teeth: al1 of which makes it nearly impossible for humans to open the sheil without breakin$ the edges (Purchon 1968: 179). The valves of the Spondylus are inequivaive, meaning that one valve is larger and overlaps the other. Also equilateral, the apeK or beak, of each valve is centred between the anterior and posterior ends of the shell (Abbott 1974: 12). The surface of the valves, or sculpture, can be either concentric or radiai and can occur sïmultaneously in the form of tidges, ribs, spines. and foliated processes. Concentric growth Iines appear on the edges parallei to the mm of the valves. Radial sculpture nms fiom the umbones. at the beak, to the margin ends of the shell, as in the ridges of spines (Abbott 1974: 12). The purpose of the spines is unclear? but they may protect the shell fiom direct attack by marine predators and by encouraging the wwth of seaweed, tubeworms, and barnacles which act as a form of camouflage (Lamprell l98W). The right valve is used to fasten the shell to the marine environment, and the left half is fiee to open and close when feeding. ïhe right half of the Spondylus is the lowermost valve, also referred to as the male half, that becomes cemented to the substratum by a secretion fiom within the mollusk ~Lampreil 1987:9; 01sson 1961; Purchon 1%8). The designation of maie and femaie haives does not necessarily relate to the mde/female dyad of Andean cosmologvt but radier is a modem designation. The natural habitat of Spondviw shells varies with each species. They [ive in sub-littoral waters and can be found at a depth of between 6 and 60 metres

dependin~CI on the species (Keen 1971:98; Marcos 1978: 103; Pauisen 1974597). While valves of both calcifer and princeps are sometimes found washed up on sho- divers were more cornrnoniy used to harvest them.

Determining the 0- of Spon@urS Materid

Because of the inability of Spondvlus to survive in the cold waters off the coast of Peru south of Cabo Blanco. between Piura and Tiunbez, it is generally accepted that SpondvIus artifacts found in Peruvian sites were obtained through interaction with the natives of the Ecuadorian coast. Similarly. the presence of Spondvlus in the Andean hi-ilands and Oriente of Ecuador was the result of interchange with the coast. However, due to the presence of a possible trading relationship with Mesoamenca (Paulsen 1977): and the suggestion by Marcos (1977) that Spondylus was traded back and fonh along this network, identification of the point of ori-pin of Spondvlus rernains an important goai. There are several methods of analysis presently in use in other parts of the world that may enable Andean archaeologists to unlock the information hidden inside the archaeological samples of Spondvius. Trace element adysis has been used successfiilly to date European varieties of Spondvhs to the time that they were hawested (Shackleton and Elderfield 1990:313). By measufing levels of strontium isotopic composition in the Sponàju1us shells the strontium Ievels were correlated to the strontium levels of Meditenanean sea water that have increased over the last 100 million years. WÏth this method Shaddeton and Eldefield(l99û) were abIe to determine whether Spondylus artifacts were carved fkom either ment or fossü species of the shelI. Locating the source of Spondvfus remallis found in the archaeological record can be dificuit. Until recentiy, no physical or chernical method was available for identifying the source of shelis ori-hating in continuous bodies of water such as the various cogsts of the Pacific Ocean. However. North American archaeologists have developed new methods of locathg the source of Busycon artifacts throue trace element analysis (Claassen and Sipann 1993 :334- 335,342). By measuring levels of magnesium present in North American shell researchers were able to distinguish shell origin between sections of continuous bodies of water- This method is based on the identification of a distinctive neochemicai profile related to the geochemistry of individual watersheds at the 5. point of entry into a larger body of water - such as an ocm. While a bay or inlet may be connected to a laqer body of water, it will still have a particular -eeochemicd signature that will be absorbeci uito the shell through the process of diagenesis. By matching the element levels in a Spondvius sheil with the unique combination of elements present in a specific semi-contained are% it mi@ become possible to iden- the point of Spodylus origin. With research of this type aiready being used to understand the role of SpondyZus in the aade activities of Europe and North Amcri- shilar procedures might determine ifSpondv2us materiai found in the Andean highlands originated in the Ecuadorïan region or even Mernorth in the West Mexican area. This wouid create a cleacer picture of Andean trade networks and their operation. The use oftnrce elernent analysis to idenm artifacts of Spondvlus is not reported for the Panamic area. However, another method of detemiining Spon@us origin for this region is commg hto use. Simpler and less technicdly complex than element analysis, it is based on measuring the distance between the primary and secondary rows of spines on the outside shell and the spine pattern (-OIafHolm1994: penonal communication). This method can on- be used for species and subspecies of princeps due to the necessity of obse~hgvariation in the groupino of the spines that are no longer present in mature culcifer sarnples. As previously mentioued: princeps unicolor from the Gulf of California and West Mexican area has rows of spines that are more widely spaced than in princeps princeps. îhe spines ofprinceps leucacantha are less crowded than princepsprznceps with variation in spine colour but with a sirnilar sculpture. it is not clear whether the spines of the different species stay at proportionately different spacing patterns throughout their growth cycb or if they channe with time. It is also not uncornmon for a shell to exhibit inconsistency in the spine sculpture pattemino (persona1 observation). The fingerprint-like sculpture pattern necessary to this method of identification can be seen to detenorate into a random pattern away fkom the centre axis of the shell. Because of the inconsistency in patternhg this technique is not conclusive. Sculpture pattern recognition is not a foolproof method of identifjring origin but neither is the use of element analysis. Both methods of research need to be expanded to include the study of Andean Spondylus. It would be necessary to look for this variation in al1 the complete Spondvlus valves found in archaeological sites before any conclusive proof could be found one way or another. ïhe drawback to this method may be that we might not be idenming species fkom different regions but dershowing that habitat ranges of these species during prehistory were more extensive. Cdour variation is dso an identifiable characteristic of different species. For example, some of the spine rows of leucacantha are white, not the ovedl coral red associated with othet shells. Identification using this method relies heavily on visual observation and is subject to the obvious problem of comparing specimens at diffierent stages of rnaturity: and shells that have been weathered through environmentai expostue. E3y these methods, new information taken fiom the sheil itseif couid improve our knowledge about shell movement and interchange. The ability to identie species fkom outside the Andean region is important to friture studies of archaeological Spondvlus specimens. ïhe inability of Spondvlus to survive the cold water of the Humboldt Current is the foundation of Spondvïus artifact study as it relates to trade and the methods of interaction that remlted in its presence in widely dispersed archaeological sites of the Andes. The identification of the sub-species is a relatively new activity for archaeoiogists. At times distinguishing shell below the species level is overlooked. In many eariy chronicles shell beads were referred to as bone, since they were considered to be fish bones. Early research into seashell materials piaced them al1 together under a generic heading: "shell". Because of these erroneous and vague identifications- some of the available information conceming Spondvlus can only be suggested by the repo- not confirmeci. Improvements in malacological identification of various shell species over the past century have increased the amounts of quantifiable data available to archaeological researchers. In the case of Andean Spondylus it is now necessary to identify the sub-species, where possible, to gain a greater understanding of trade relationships. CHAPTER THREE

Defining mullu is more problematic than descniing Spondylus. ïhe list of characteristics includes a group of abstract concepts that Vary according to geographic location and use. Chroniclers used the term mullu in a purely physicd way when refemng to sea shells of a white or coral colour (Gonçalez Holguin 1608; Santo Thomas 1560). By understanding the concept of muflu, an understanding of its place within the socio-cultural bel ief sy stems and ritual of the Andes, and the trade networks that handled its distribution, cm be developed. As translated by Diego Gonçalez Holguin (1608), rnullu is defined as

"concha colorado de la mar chaquira, O coral de la tiem," which refers to beads made of red sea shells, or coral of this land. The related term mollo (sic) also refers to coral- or pearls- as translated by Domingo de Santo Thomas (1560), confirming an ambiguity of definition that could lead to alternative sea shelis such as the Pinctada mazutianica or Pteriu stema, more commonly known as the pearl oyster, taking their place alongside Spondylus as mullu (Mester 1990:Zl). The coral colour of some parts of the Spondylus shell, when seen as finished beads, may be responsible for this definition. Another word, with a similar pronunciation - muyu - meaning "round", may be related through the roundness of the beads or necklaces on which they were strung (see below). The suggestion that mullu couid actualiy refer to a coral is disputed by Petersen (1970:25) who identifies most of the coral found in Andean museum exhibits as materiai made fiom the spines of Spondylus. Some confusion may have been created by the use of the pidform "cordes" which aiso referred to bracelets of coral material (Diccionario De Autoridudes 1976 [1726]:589). If the coral material was in fact Spondvius. then the bracelets may have been made of SpondvIus? caiied mu12u1 but referred to as corales by the Spanish. This is an example of the semantic problems that hinder the definition of mullu.

Ortbogrripîüc Variation

While the orthographie variations of Quechua words create confusion as to the intended meaning behind words used in the chronicles, the -pfammatical use of those words as roots for other words and actions can provide the key to unlockino the lost information. In the chronicles, Spondvlus is known interchan-biy as niullu, nioh and mullo. ALI three tenns appear in the toponyms, phrases and chronicles of both Pem and Ecuador. However? there appears to be no set mle as to where these derivations of the tenn were used. Confusion about the specific meaning of each of these words, and whether their use was resmcted to a direct translation for SpondvIus. is apparent in the definitions of each variation and derivative forms- Problems of pronunciatio~orthopphyZ and literacy were experienced by the Spanish as they attempted ta put a written fonn to an unwritten language - runa simi. The written fonn of Quechua was filtered through a foreign linguistic system that reflected what the writers thought they heard - not what they had heard (Mannheim 199 1: 127). Even the indigenous chroniclers faced a pro blem of interpreting their own meanings and linguistic signs into another language (Harrison 1982:68). One reason it is not possible to translate Quechua phrases literally is due to the differences in language patterns and word signification based on pronunciation of similady speiied words (Mannheim 199 1: 128- 139. By Iowering the sound of a final syllable Ït is possible to change the entire meaning of a word. îhe issue is Mercomplicated by the absence of a sin@ fonn of the Quechua langmge. Dialectic variations of Quechua were a reflection of ethnic identity and differed fiom vailey to val- (Cobo 1979:39; Harrison 1989:35; Mannheim 199150). This problem of spelling variation in the Spanish interpretation of the Quechua languwe mi@t be explained by re@onal differences in pronunciation- but the problem is exacerbated merwith the inconsistent use by researchers of the various spellings- The Spanish inability to distinguish Quechua vowels was also compounded by the lack of a written record with which they couid correlate spellings wîth pronunciations; a problem that modem researchers do not have to face. Both Hyslop (1 984:248) and Rostwomwski (1 977: 176) substitute the more recent orthographie spelling ofmwfor mullu when describing Spondvlus. If muyu is a variant spelling representing mullu then it might explain why toponyms such as Mullupungu appear in Ecuador and Muyupongu appear in Peru. The lack of toponps in Peru with the mullu prefix is puzzling as the geographical separation in use appears to follow modem political boundaries. However, rnuyuZ defmed by early dictionaries as meanin3 round, does not directly relate to the definition of mullu. The modem use of the word nup dates more ta the action of winding or circulating (Lastra l968:9 1; Parker 1969:164). ïhe only possible connection of rnullu to roundness is its association widi events of a cyclical nature such as birth, rebira and regeneration. It may be that there is no identifiable association between the words mullu and mu_vuZand as such the Iack of mullu toponyms in Pem can become the subject of odiet investigations. Conversely, it is possible that ail of these definitions relate in a complex of meanings. In another example of mullu vm-ation- Hyslop (1984:35) discusses the region of Mullepungu !sic) southwest of Tomebamba refemng to it as a

Spondvius door dong a Spondvlus trade route. In this case, 'd and 'O' appear interchangeable in the orthogaphy of tnullu indica~gthat Mdlepungu could be similar to Mollepunw But molle is a type of tree th& when bwnagives off an odour similar to incenset acts like a laxative for animals when the ground-up bark is ingested, and whose red bemes were used for the makino of chicha (.Cordera 19 11 :40 Olaf Holm 1994: personal communication)- none of which indicate a connection to the definitions of mullu. If there is a connection, then perhaps it is found in the relationship of these items as they might have been associated with inullu during sacrificial rituais. Further anaiysis of these tenns and the uses to which the objects were put may explain the etymoiogical and metaphoricai relationships of these words.

Things That Are Mulla

ïhe identification of mullu has at times been hampered by an uncertainty over materials used and theu categorization. A distinction is made that mullu couid be Stone, bone, or shell. Coloured travertine, a buff-coloured porous minerai forrned by the deposition of calcium carbonate in hot sprinos, such as those near Cuenca in Azuay, is difficdt to distinguish in the archaeologicai record fiom Spondylus shell material (Bushnell 195 1). The travertine hardens upon contact with the air and can then be made into beads similar to those of SpondvItlsS Spanish chroniciers wodd have referred to the materid as mullu based on its similarity in colour and materiai. At thes Spondylus beads have been identified as bone, based on its eariy detemination as fish bone materid simply because it came nom the sea (Olaf Hohpemnai communication). A description of chaquira - beads made of red and whne Spondylw sheil - by Cieza de Leh (1962:99) describod them as "long strands of fine boae bds?white and red."3. Agathe mistaken identifications included in hktorical accounts makes the determination of materid difficdt. Some mullu cornes in the form of necklaces orjewellry cdled mulla Often the mdIo is made of pearls and precious Stones (GuamanPoma L980:306?916). Because some of these objects have no Spondvlus materiai in them, it is difflcuit to determine whether it is the jewellry and its use that is referred to as niullu, or if the altemate materiais used in its manufacture can also be considered as mullu. In Ecuado- small necklaces of green material are currently known as muilo. In the Museo del Banco Central in Cuenca a miniature necklace made of stone from the site of Pumapungo is mullo by Caeen materiai listed as a Max Uhle !1994: personal observation). Uhie (1922:208) also describes the artifacts hm Chinguilanchi near Loja in this way:

Al1 of the subsoil of the floor contained nwnerous shelis of Spondylus QktoNrn) offerings inlaid with green, red?white 'muilus' of stone and shell. In various parts of the subsoil we also found ~roups dfierent of mullu; below the C of classes unattached same al- were piled a group of 40: more or 1- Spondvlus shells-4

-- - 3 "...Traen atados grandes ramales de cuentas de hueso menudas, blancas y coloradas?que ilaman chaquira' (translation by Salomon 1987b:66).

4-"..îodo el subsuelo del pisa contenia numerosas conchas de Spondylus (pictorum) votivas rcllcndas con mullus vcrdcs, cotorados, blancos, dc picdra y concha, y objectos de otra clases. En varias partes del subsuelo se encontraron In this case, Uhle is drawng a distmction between Spondvlus sheli, mullu of Stoneo and the offerings filled with _pe- red and white inlays. 'The description of difEerent classes of mullu and coioured mullos may have been created by Uhle as descriptive terms which do not reflect pre-Columbian native - and as such are irrelevant to this discussion. The identiikatïon of red and white mullu can be related to the red and white of the Spondym shelf, but the use of green muilus brinos a new dimension to the definition. The problem of defining mullu increases when the question of mullu material is combined with questions of coiour association,

Technicoiour Mu&: Colour In The Chronicles

The use of colour spbolism in the Andes assigns various objects to specific symbolic complexes which form dyadic oppositions, such as the celestial and terrestriai complexes. The placement of Spon&lus in the terrestrial ymbolic complex of eh,irrigation wate- and agicultural femlity, rem on the colour association of the shell's red rim to red eanh and pottery clay (Davidson 198 1; Mester 1989: 165). Ann Mester (1 989:l62; 1990) piaces the pearl oyster in the celestial symbolic complq but considers both materials ta be nullu. Both these complexes are related to mascuhity and femininity - mullu can belong to both of them. The celestid complex is masculine and offerings of Spondvlus would bring thunder: li&tning and torrentid rains: whüe the female terrestrial complex is connecteci to springs. lw soft rains and the sea (Davidson 198 1:8 1).

------tambien grupos de las diferentes clases de muilus, en estado libre; debajo del altar misrno, un ppode, m8s O menos, 40 conchas de Spondytus amontonadas." These associations are based on coiour assipments and a cornparison of drawiags by Gu- Poma that suppody iLiustrate differences in object status between Spondvlus representing agricuitural femlity, warfare and the low status non-Inca peoples, and pearl oyster representino hi@ st;ttus, beauty, and moral excellence (Mester 1989:160). However? the Iitter referred to as the "quispi runpa", identified by Mester as the "Litter of pearIsl'. couid be identified as the litter of qstals. Guaman Poma refers to it as a litter of precious stones 5 (1980 [l583- 16 151: 304-305/33 1[333]) and descri'bes quispi as pearls, precious Stones, such as emeralds or cqstal, and neckiaces of mal10 (1980 [1583- 16 151: 306/332[334]). In this case mullu is separated nom pearls and other precious stones in the defmition of quispi. Whether or not Mester is right in her assessrnent of mothersf-pearl as an object that can be seen to represent hi@ status, beauty, and moral excellence based on the wrïtings and drawings of Guaman Poma requires meranalysis. mever: I think Mester (1 990) is correct in suggesting that muilu refers to other colours and shells: and that mudreperla. as a coloumil object associated with the colour symbolism of the Andes (Mester 1989: 162- 1651, can aiso be classifieci as a nullu. The colour symbolism used by Mester for her argument is based on Martin de Munia (1987 [1590]:420) who described the ground sea shells used in sacrificial offerings as paucur mollo and yahuur mollo 7. The shells have different names according to their colour and were offered for different purposes?

5 "ANDAS DEL INGA QVISPI RANPA".

6 "...Perlasdel Ynga y de sus prencipales que son piedras preciosas. Se llamauan quispe pcrlas, picdras prcciosas mina. cuychi. uaccri y colarcs SC llamaua muh. Estos dichas cosas auia en tienpo dei Ynga en este rreyno".

7 "...Otras veces ofTecian polvos de almejas de la mas, molidas, que dicen paucar mollo y y ahuar mollo,.. ." an identification dso made by Acosta ( 1962 [1590]:247) and Cobo 9( 1990:1 17). This difference in shelI types and colour classification 1s also referred to by Cristobal Molina (1 989 [1575j: 68) when desctibing the different muhs as red and yellow seashells 1% Mester (1990:213) defmes paucar as light and bris& although respiendent would be more appropriate? therefore indicating that the yellowish shhe of mother-of-pearl is the paucar mullu, while yahwr. or blood&): refers to the deep red Spondjdus shell. ïhe association of yahuar to red Spondvlus may be ap& but Munis might be associating paucar mullu with the white part of the Spondjlus shell when poiished to a hi@ gloss. This interpretation may be more appropriate considering Munia [ 1987 11 S90]:42O) was aware of chaquira of mullu that contain both red and white beads of the shell. Rather than a symbolic dennition for types of muIIuT perhaps paucar and yahuar were simply desipatioas for the two different colours - white and red - and the quality of their lustre- As the rnaterial and colour associations of mullu become clearer the definition becomes broder, to the point that objects are not just mu& but muilu-like. Molina ( 1989 [1575]: 133) describes a cerernony to bnng about good health where several varieties and colours of rnaize, includhg some with red and yellow stipesJ are gathered with coloured mollo mollo that is known as

8 " .. -[conch llab mollo] Tienen diferentes nombres segh la color, y asi simen a di ferentes efectos" .

9 "...Accordhg to the color, the shells were offered for different purposes, somctimcs wholc, othcr timcs ground into powdcr, othcr timcs brokcn into pieces, and in addition, some fimeswere made fiom the powder" [translation by Roland Hamilton].

10 Y.- y conchas de la mat que llaman mullo, colorado y amarillo, hechas a mancra dc maiz, ..."

11 .".-y maiz entre vetada de colorado y arnarilio que llaman çunqaraf'. ~vma_vmc~~uimollo,~2 and then -muad tosether and blown on the -rmacas 13. The term~vm~imranrdlo probably refers to al1 kinds of mdl~while the use of the reduplicative noun mollo mollo infers Iikeness or similarity to mollo (Parker 1%9:100). It is obvious fiom the above that not ail mulh is made of Spondvlus materiair nor is it al1 of the same coIour. From Cobo (1990:69), the list of shrines and guucas on the road to Antisuyu includes a guacu cailed Lampapuquiu to which was sacrificed "shells of two colours, yellow and redW The use of red and yellow colom within the context of ritual or sacnficial offerinp has su~ved to the present day. Modem ritual traditions in Pem use rnesar or chann bundes that include red and yellow maize (GBord and Hoggarth 197654) for ceremonies like those mentioned above by Molina. In one of the myths of Huarochiriz the god Paria Caca sent red and yellow rain ls ont0 HuailalIo Caruincho, which was dammeci up by one of Paria Caca's alter egos called Llacsa Churapa creating the lake known as Md10 Cocha (Avila 1975 [1 S98]:52; Salmon and Urioste 199 1:68). in this case the rains are the same colour as the types of corn used in sacrifice. Mullu colour association is Mer augmented by the addition of bluish-green, or turquoise, to that of red and

'.--yotras conchas de la mar que llaman ellos mollo mollo de todas las colores quc pucdcn avcr, quc llaman ymaymana mollo;"

13 "...A todas las guacas y uilcas quatro partidas desta tierra, agaellos y antcpasados mios, rccciid cstc sacrificio doquicra quc cstais y dadmc salud."

14 "[AN-9:1] la primera se decia, lampapuquiu, era una fiente que estaua en [tachado: vndamarca] Vndamacha [sic; dcbc dccir Andamachal sacrificauanlc conchas de dos coiores, amarillas, y coloradas" (fkorn Rowe l979:38).

15 "...desde chco direcciones hizo caer torrentes de Iluvia; esa Iluvia era amarilla y roja;" (Avila 1975 [1598]:52). yelIow? and green. Llacsa Churapa has the same fim name as the "bright

Imeenish-blue colour that blew fiom the mouth of Mkca Uisa iike smoke as he spoke"I6 (Avila 1975 [1598]:52; Salomon and Unoste 199 1: 115) after eating mullu. The original Quechua term used in the Humochin' Manuscript @alornon and Urioste l991:214) for describing this colouml smoke is IZacçu (Zlacsqi, a term that also refers to smelted or alloyed metai matenals such as copper or bronze (Gonçaiez Hol-pin (1 608) 1952:207). As a colour Zklcsa is described as a green powder or Stone, Iike copper oxide (Amiaga i968[1621]:46),similar in colour to turquoise. If llucsa is a by-product of mullu consumption then there could be an association between turquoise coloured objects and mullu- During sacrificial ceremonies the person in charge of the huaca would ask for offerings of mulZu and Ilacsa (Arriaga 1968[1621]:46) indicating a rinial value for llacscl. The connection between Ilucsa and nrullu forms a foundation for the suggestion that the colour green, or greenish-blue? is also associated with mullu in a very strong symbolic way. Based on the colour association and the use of the word llacsa to describe both the colour and materials, then turquoise material can also be associated with mullu. The connection between Ilocsu and turquoise may be based on colour or the fact that turquoise as a phosphate of aluminurn includes traces of copper that -sive it its bluish-gren hue (Clark l986:69; Grieder l988:87). The material called llacsa was important enough in the Late Horizon for the creation of a position known as the Zlacsa camayoc: a person who worked airquoise and stones from the sea (Rostworowski 1979335). Copper is one source known to produce the powdered oxides known as llacsa. But is it the -oreen-blue colour or a specific

16 '.-.Ymientras hablaba, brotaba de su boca el diento y una especie de vapor azulado" .(Avila 1975 [1598]:52) substance that is referred to as llacsa? Bluish-Pen smoke, bluish-greenv turquoise, and bluish-geen copper oxide al1 can be defined as Ilocsa- The association of tuquoise and muI2u goes deeper than just the colour of exhaled nrullu gas. Turquoise and Spondvlus appear together in many archaeologïcal contexts suggesting a ntud comection (see Appendix A). Near Cuzco collections of 40 turquoise figurines each were found at the Middle Horizon site of Pikillacta in context widi Spondylfls princeps, Strombus shell and a bronze bar i7 (Cook 1992:344). At Marcahuamachuco, dm Middle Horizon. turquoise materid medin the shape of srnail Spondyius shells was found in association with reai SpondvIus ('ïherea Topic 1994: personal communication). ïhis association appears again at Huacho. north of Chancay, where a Spondylus bivalve was recovered with a small pnstone id01 and three small stones inside (Cook 1992:359). URfortunately the Huacho Spondylus is not dated to any specific time perïod. However. the three small stones were described by local inhabitants as "the fm Iima bean", "wheat". and "chili pepper", which would indicate a post-conquest interpretation. The association of the colo ut green, turquoise-like material and muiiu is becoming stronger and better defined through archaeolo@cai contexts and ethnopphic research. Modem ethnographers have identified the importance of the colour green to the Andean people through its association with anceston and souk of the dead. Like muilu which signifies rebirth and regeneration, the word rnaliqui describes both the green branches of the angoripa tree and dead ancestors, si-ifying renewai and continuation (lsbell 1978:147). Also, on the first of

17 The 'bronze' bars referred to by Cook have been identified by Juan Lmea [1%0:59-94].as copper scepters known as yawi - the Quechua word for the Inca royal st& Yauri is an Aymara word for copper, and the presence of the staff possibly indicates royal statu for the fimes. Novem- duruig Kawsasqan~his~the souis of the dead return to visit their relatives. Those who have seen the dead souls' ethereal presence? describe them as "pretty and pn"(Allen 1988:164). The use of green objects and the coiour -nreen as a description for important concepts requires more study thaa is possible in this thesis? but the connection between these events and mullu is quite clear.

Another entry in the dictionary of Gonçaiez Holguin (1608:249) lists the words "niuiluysi~ipucaysinti rnull-1 virpa", and translates them as "el de los Iabios colorados hermosos." When transtated the words hold shilar meanings: rnulluysinti~having a mullu mouth; pucaysimi? having a red mouth; and rnuiluy virpa, having mullu lips. With the word simi relating to the mouth or speech? and by eliminating the Quechua words virpa for labios and puca for colorados, we are left with on- mullu to correlate with hennosos. The translation of Gonçalez Holpin's definition to English wouid be the person with the beautifil, or lovely red lips- Whether it can be translated directly in this way is questionable. ït is possible that the words refer to a more absûact concept that is not yet clear. The key to this discussion may centre on the presence of the letter y as a suffix to mullu and puca. There are several possible reasons and actions associated with the use of y in modem Quechua grammar and syntax that might explain its purpose in this situation (Lastra 1%8; Mannheim 1991; Parker 1969; Sola 1967). In some cases the addition of y as an allomorph of the niy sutnx indicates a first person relationship to the nouni such as: wasi - house, and wev - my house. It can create a noun that hctions as an adverb refem-ng to an abstract actio~such as: lamka - to wo& and Iamhy- working, or it can ïndicate a direct command when added to a verb stem- With these options available, the role of the word rnullu can be regarded as that of noun. verb, adverb, or command. UnforauiateIy. determinhg the intendeci form is dificult wîthout the accompanying vocal innection that would idenethe proper meanhg All of the above phrases refer to red Iips or red mouth through either the inclusion of the Quechua quivalents or because mullu is described as having red lips. There is a similarïty between the use of mullu in this sense and the red rnargïns that encircle the edges or lips of the Spondvlus shell. The physical appearance of SpondvIus as an object with red lips is one characteristic that relates it to female genitalia The Spondvlus shell has been described as a representation of the femaie genitalia representing a mythical vulva that was protected by its intertwineâ spines (Marcos 1986: 198). An understanding of the use of mullu in the above definitions may be dependent upon understandin8 this other gender related association. A more recent definition for rnulIz~simz,without the y- is " mujer de lab ios muy rojos y hermosos" (Lam 1978:146). While the definition may be based on that of Gonçaiez Hol-in there is a change in gender reference that is confusing. The lack of ay could be the result of syllable-ka1 weakening over time where the uy diphthong was simplified to a u in its position at the end of a word (Mannheim 1991 -216). With simi (mouth) being the subject &ara 1978:UO): the adjectives relating to rojos (red), hennosos (beauh.), and mer(gender), are al1 included in the reference to mullu. When translated into woman with beautifid red lips, this becomes Cnender specific context symbolic. ïhe Gonçalez Holguin definition is not specific about gender using el, which can mean either he or the person, instead of mujer (John Topic 1994: pemnal communication). The question then becomes? "1s this a reference to a female who has unusually love- lips, or does the phrase connect mllu to the femaie genitai area?" As an important rinial obje* the value of mullu may be partiaIly related to its resembiance to the vuiva as it relates to birth and the brin-dg of new lifie. It is clear tbat the use of the word mullu in these phnises does not directly relate to Sp~ndvlas~but instead to a concepaial representation of the shell. This is a Merexample of how mullu can not be transiated directiy into Spondvius shell or chaquira and, as such, indicates again the many meanings and associations available to the word- The femde metaphor continues into an examination of the basic structure of oppositions in Andean cosmology. The appearance of Spondvhs in the dyadic structure of IeWright, male/female is based in part on the physical characteristics of the shell (Burger and Salazar-Bwger 1993; Cordy-Collins 1978; Paulsen 1974). This duality is represented through the separation of left vaive/right valve and the gender identification of the male valve/female valve: an assi-ment of attributes that can sali be found today among the Desana Indians of the northwest Amazon (Reichel-Dolmatoff 1971 : 167-71 ). Our own dichotomy of a physical sheiI/concepnialized shell illustrates both the concrete and abstract traits of Spondvl~s/mulIu- Spondvlus representations at Chavin de Huanuir are well documented and appear in various forms such as on the Tel10 Obelisk (Figure 12), although none as important as in the carving of the SrnilM God dated to phase D*at about 400 BC (Bur~er- and Salazar-Burger 1993:98; also see Figure 13). The carving is important not just because of the presence of Spondvius~but because of its association with a Strombus shell and the dyadic structure that the two shells represent. The gender identification of mullu as a female symbol of disorder, destruction and regeneration is iiiustrated by its placement in the Smiiing God's lefi handowhüe in the rÏ@t hand is a S~ombusshell. the male symbol of order and structure &athrapSCollier and Chancira 197559). A simiiar presentation of these malelfemale associations can be seen on the Tel10 Obelisk (Lathrap l985%93O). Spondylus was fond not only in the represeneations at Chavin during the Ear- Horizoq but was used by elites in ceremonid centres and burials in both the coastal and sierra re~ionsof the Andes (Norton 1986: 135). Textiles fkom the Carhua site south of Paracas show Chavin related motifs and designs that included Spondvius shelis. Specimens that show representations of staff gods and goddesses exhibit characteristics of Chavùi iconography. Staff pdC- 1 (Fig. 14) holds a pouch of Spondvlus shells, with a SpodyZus shell wing on his back, while staff god C-4 (Fig. 15) is shown with staffs, headband, and waistband finials that al1 end in Spondvius shells (Wallace 1991a:75-76). The gender identification of supematurals (see Lyon 1978. Stone L983) verifies these images as male gods with S'ndvIus. it could be expected that female supernaturals wouid be depicted holding the shells but that is not the case. It is apparent that Spondvlus as a female metaphor and staff goddesses wiU have to be tied together through other means. At Kuntur Wasi in the northem highiands of Pem tombs were excavated fiom the Kuntur Wasi phase (760 B.C.460 B.C.) that appear to illustrate the maie/Strombur and femaielSpond-lzu relationship. Excavations of Tomb I included a 50 - 60 year old male dong with Strombus artifats - three aumpets. two small Strombus beads, one Strombus plate and one Sfrombus pendan&while Tomb 4 ,a 60+ fernale? contained 849 beads of red Spondylus in various shapes and sizes (Kato 1993:î 17-22}. The male/female idemincation with Strornbus and Spondylus as uidicated in the iconography of Cha.was put into pmctïceat the site of Kuntur Wasi. This is not to say that d the maie tombs uncovered at Kuntur Wasi incIuded Strombur &&as; Tombs 2 and 3 did not However, the fact that the femaie and male buriais excluded the other type of shell is sipificant in this case, This relationship between females and mulh continued throughout succeeding time periods until the end of the Late Horizon. At Inga Pirca in Ecuador Spondvhs fktgments were found in the area now known as the Temple of the Moon (PiMoma) where seventeen female skeletons were entombed. While the site was eventualiy used by the Inc* the Temple of the Moon is located in the section built by their predece~sors~the Catiari. Archaeolo@cal evidence of this nature is used to corroborate the chronicle infornation concerning the feminine association of Spondvlus with the male/female dyad. However, it should be noted that Spondvlus was not used exclusively by women? and in other contexts appeared as an indicator of other associations. The relationship of mullu to water? water sacrifice, and female characteristics is an important one. The placement of water, the moon, and wornen in the leftkight dyadic structure of Andean cosmology can be seen in the diagram drawn ôy Santacruz Pachacuti Yamqui to illustrate the cosmogram found on the wall of the Coricancha in Cuzco (Harri-son 1982:75.1989:66-69;Silverblatt 198 1: 164- 169; see Fig. 16). The presence of water, the moon, and ancestors on the female side indicates a complex of identifiable associations. The leWright theme continues throughout the Andes and can be seen in the drawing of Guaman Poma. Action is separated into upperAower, IefWght, male/femaie, where the male complex traits occur on the ri@& and female activaies appear on the le& unless they are dominant to the theme of the picture (Adorno 1982). These associations make it clear that mullu was considered to be pan of the femaie symbolic system of associations. Associating the definition of mulluysimi. pucaysimz. mulluy virpa as the woman widi beautifid red lips with the deep red Iips of S'&lus princeps and the belief that Spondvlus represents a "mythïcal vulvaw(Marcos 1986:ï98X presents an interesthg possibility. Any physical manifestations of this beiief are difficuit to locate. If the Spondvius shell with its bright red margïns can be seen as lips, and its long le@ spines can be seen as interlocking tee& then what we are looking for is a female fi-mire with a vagina dentata. Such a figure can be found in the visuai presentation of female staff -goddesses on Carhua textiles. The ar-ment for the presence of staff goddesses as opposed to staff gods has been made convincingiy by Pamcia Lyon (1978) who recognized the vagina dentata as one of the characteristics common to the female deities (Figues l7J8). ïhe vagino dentata is not unique to the Andes. and in fact appears around the world in the mythologies of many cultures (Thompson 1955:833). What is interesting about the vagina dentata and mullu is the similarity by definition and pbolic representation of the two objects. The vagina dentata motif represents women who have teeth in their vaninascc (Lyon 1978%): the sisnificance of which is not known in this instance. Visualizing the spiny projections as teeth could make the Spondvhs shell appear as a disassociated vagina dentatu- When the lexical meanings of mullu as it related to red lips and the association with female genitdia are combined, the definitions mentioned above take on a new meaning. 1 suggest that with the female associations, ideolo@cal implications and genital references to mullu already discussed. that SpondvZus princeps (the thorny opter) is a living representation of the vagina dentuta. if the motif accurately represents a Spondvlus shell, then the teeth senre a pmtective purpose rather than one of eating and if it is a Spondyïus shell then perhaps it is located in this position because the SpondvZus shell was a symboi of rebinh and femlity. Whether the vagino dentata symbolism was based on the Spondjdus shell and the concept of mulk or whether Spondylus was imbued with the femaie characteristk of rebirth and femlity due to its sirnilar physical appearance to the vagina dentuta is unknown. Were Spondvlus and its metaphoric portrayals responsible for the depiction of the vagina dentatu? or did the visual representation create another role for the aiready gendered sheil to pomay? We shouid remember that vagina dentata is a non-native Latin term used to describe this rnythologïcal imagery. To use it in the Andean conte- does not indicate that the Andeans were using the same concept of vagina dentata. [t rnight be more appropriate to consider that vagina dentata is a usefiil term for sornethîng that is similar by definitio~but which is more correct@refemng to a Spondvius vagina. 1 would aiso sugsea that the words mullu_vszmi,pucaysimi and mulluy virpa are ail references to this concept of a Spondylus vagina and represent the verbalization of the image in everyday life. This connection is further strengthened when conside~gthe second- image- of the breasts of the staff goddess as eyes and the Spondjdus vagina as the lips and mouth. mis concept in itself heips differentiate the associations attrïbuted to the shell even though it is referred to by the ail encompassing terni mullu. Mullu in the chaquira fom could not be considered as a representation of female genitaiia as the vagina dentata, or vice versa. The complete bivalve is required to accomplish this connection. Ruo. Lakes and Sprigs: Water Associations

The association between nullu the -and sacrificiai offe~gis related by several chroniclers. Acosta (1962 [1590]:247) describes the offerïng of

"conchas de la martvzthaî the natives cd1 mollo, to the fountains and springs 18. The shells were said by Acosta and Mh(1987 [1590]:422) to be daughters of the sea, also known as Marnacocha the mother of ail the waters. A similar version of this theme is presented by Bernabé Cobo (1990 [1653]: 1 17) who identifies the springs and fountains as the ciaughters of the sea, not the Spondylusl? Whether or not it is the shells or the springs that are the daughters of the sea is an interesting point. SpondvIus cornes fiom the sea and could be seen as a child boni of the sea. This could be an illustration of the comection between Spondylus and the fernale genitalia ifthe shells are king referred to as daughters of the sea based on theu ability to qmbolically bear Me-giving water. It may be that both descriptions are appropriate as both the shells and the springs are important to the provision of water and as such could both be daughters of the sea. According to Kamen (1926:384) the shells wodd contain some of the "water-power of the great ocean" as daughters of the sea, and as such could transfer some of the ability to always supply water through their sacrifice to the springs. ïhe reasoning behind this assumption can be found once again in

18 " ...Iten también sacrificaban u ofrecian conchas de la mar, que Ilamaban mollo, y oftecianlas a las fientes y manantiaies, diciendo que las conchas eran hijas de la m- madre de todas las aguas".

19 " ...These Indians were dm accustameci to sacrifice seashells, especially when thcy madc ofScrings to thc sp~gs.Thcy said that this was a vay appropriatc sacrifice because the springs are the daughters of the sea, which is the mother of ail waters" [translation by Roland Hamilton]. Cobors (1990:60) kt of guacas. On the seventh ceque of the road to Chinchayswu was a spring caiied Cailancapuquiu to which was "offered shells so that it would always flow20." The suggestÏon that riluilu could be responsibie for the production of water appears in the Koniraya myth hmthe Huarochiri Mantlscript @alornon and Urioste 199 1:66). A batde took place between Paria Caca and Huallallo Caruincho who lived on the Ida de Mar in MuUo Cocha a lake that was uihabited by mollusks (Tel10 and Mejia Xesspe 1979:38). The mollusks were expelleci from the lake by Paria Caca in reprisai for flooding the earth with the waters of the sea and Mullu Cocha (Teilo and Mejia Xesspe 1979:38). The responsibility for the flooding was placed on the sea mollusks (Spondvlus) imbuing them with the ability to control or provide water and rain. The lake and its name have swived into the present day as one toponym that does use the II radier than y spelling for the place name. Muilucocha is located in Peru about 30 km east-northeast of Huarochiri at 4.300 m.a.s.l., a highland location that would not have been home to shells fiom the sea It is likely that the name was associated with this location either because the myth required the presence of mullu to account for its sacrifice to the god who produced r- or that quantities of Spon&ius shell had been sacrificed at this location. An underwater survey of highland lakes such as this mi@ be usefid in determinhg if sacrificed Spondvlus or fossilized shell is present. The relationship between women, water sources, mu- and the daughten of the sea is made dl the more complex by the suggestion that a relationship exists between women uruiating and the creation of spring and other water

------" [CH-7:6] la sesta se decia, callancapuquiu, es el manantid de Ticutica [sic; debe decir Ticatica], al qua1 ofhian conchas por que siempre manase (Rowe 1979-24). sources (Silverblatt 1981:33). If women could create water sources Wou& urination, and ifmullu and the water sources are "daughters of the sea" - with nullu also representing the femde geoital area (vagina dentata) - then the association becomes stronger. An alternative explmation for the reiationship between Monand the creation of sprhgs might be that it is the amniotic fluid emitteci prior to chrldbirth, and not urine, that is sem as causing the creation of water sources (Joan Vastokas 1995: personal communication). Womea, water, and mullu are ali related to each other, either alone or together. Mullu is aiso indirectly Iinked to the moon thmugh its association with women and water in a fernale complex of characteristics that appear in the cosmagram from the Coricancha in Cuzco. The most powerfùl element of the complex is the moon. It is the moon that appears on a qclical basis affecting and seasons of raino both which are associated with mullu sacrifice? but more importantly its cycles are similar to monthly mensmial cycles. The lunar effect on tides might al so affect the activity of mullu harvesting (Daniel Sandweiss persona1 communication: 1995X presenting specific time periods when the operation was more easily conducted.

Food For The Gods, Food For ThePtople

An alternative use for mullu cornes again fkom the myths of Huarochiri (Avila 1975 [lS98]: 105; Salomon and Urioste 199 1:67-8,116), where the idea that it was consumeci by the gods first appears. The meaning of mu110 is translated into Spondvlus as it refers to the activity of bringing (rnullon), giwig (rnulloyquztac~)~and eating (mullocta) the thorny oyster shell. When Tupay YngaCI Yupanqui summoned the huacas to Cuzco to aid him in his attempts to quash provincial rebellions, it was Maca Ui- son of Paria Cmwho ageed to provide heip. With the job completed the Inca provided Maca Uisa with celebration and food to which Maca Uisa replieci. "1am not in the habit of eating stutr like this. Bring me some thom. oyster shells, [eath] hem all at once, making them crunch with a Cap Cap sotmd." (Salomon and Urioste 199 1: 1 16). From die transcriptions of this myth it is the sbell that is supposedly being consumeci not the meat of the moUusk. ïhe suggestion that Spondylus meat is harmful to humans at certain times of the yeq capable of causing death (Davidson 198 1) may conmbute to this view that it is oniy the gods who are capable of eating it at any the of year, but there is no indication in the myths that anything more than the sheti is king used. However, the Spondyius meat is quite edible when cooked (personal observation) and could have figured in the diets of coastal groups in Ecuador where it could be bhly obtallied. It is this myth that has conmbuted to the idea that Spo~vlusis the food of the gods. However, there is no indication in the Quechua version that mullo is referring to Spondyl~s22~rather that the god is simply asking for mu- an unidentified material. to be brought. There is no indication as to the morpholo~icalsubstance of mullo apart fkom the act of grindine of the teeth which infiers a toughness to the material. In many cases both modem researchers

- - 21 " ... Yo no me aliment0 de estas cosas. Manda que me traigan mullo." Y cuando le trajemn el mullo 10 devoro al instante: " j cap, cap! ". rechinaban sus dientes, mientras masticaban(Avila 1975 [1598]: 105)

22. [299] chaysi micoy yaya nispa carachiptin ca manam ilocaca cay ynacta micoccho cani muiiocta apamuy nispas rnadarcan chay si mdlocta coptinca cap cap IIichispa tui lla mico m...(Salomon and Urioste 1991 2 15). and chroniclers have resnicted the definition of mullu to a specific meaning Spondvlus shek and help propagate the idea through their own work This is an example of the different uses and associations of the tenn mullu which have spread over the Andean area Associaîions that occur in one region and time do not necessarily correlate with those of another region and tirne. If they di& then the consumption of muliu by a gaiof Huafochiri could be seen as the performance of cunniLïngus on the vagina dentaro of the staff goddess hm Carhua. Howevert there is no evidence avaiiable that would support this suggestion and no indication that the associations were shared. Other references indicate that mullu might aiso have been used as food of the people- or at Least as a condiment for the people's food. Excerpts f%m chronicles that follow include references to ~cumul2u~an herb used in cooking. ïhe definition of paca as it refers to sacredaess in association with a rituai object such as mullu is made al1 the more interesthg by this association with plant material. The identification of rnullu as an herb is confinned by Gonçalez Holguïn (1952 [I608]: 165) who dekes huaccarnullu as a "cierta yerua de corner", but the type of herb is not identified. The question is: why would an herb be referred to as mallu if it is not Spo~ltcsshell? This is another example of mullu king used as a reference to objects not made from shell. An initial reaction would be to idente guacamullu as guacamole but there appears to be no relationship. Avocado, or palta, was known by the Spanish when they arrived in Pem and it was not prepared in the manner as described in the following accounts. While there are similarities between guucamullu and guacamole as a green coloured foodstuff? its identification as guacamole appears to be a rdherring. The fïrst reference comes hmthe Relacih y àescn'pciion de la ciudad de Lmnear the southeni border of Ecuador written in 1571 by Juan Salinas Loyola (1 57 1: 2%). In it he desmï'bes the use of various vegeîabies and herbs? including one called guacmol?o that is eaten with or without vinegar and cooked with meat 23 . A similar description is given by Sancho Paz Ponce de Leon (1582:238) in the Relaczdn y descripciOn de lm pueblos del purtïdo de Otuvaio concerning the area north ofQuito in which he also refen to the heib called guaca-mullos24. In both cases the reference is to an herb called 8~~fcc1mulluor htlc~camullo, and in the Saluias Loyola account it appears to be describeci as green (verde) and prepared by boiling. Those plants whose leaves were boiled and eaten are referred to by Rowe (l%3 [l9461:2 16) as yoyo? but in die Reluciones geogrcrficas de Indias (Fomee 1965:18 and 22) they are called pyus 25. An alternative meaning for verde could be unripe and not an indication of colour (John Topic 1994: personal communication), but the relationship between nrullu and the colour green has already been discussed and could be relevant in this case. A third reference is found in Guaman Poma's Nueva Corbnica Y Buen Gobierno (1980:/338[340]) where he describes a list of items? including uacamullo, given as a token of payment 26. The only indication of what

- 23 "...La yerba que rnh sime alla de hortaliza que tenian los naturales, Ilhase -nuacamollo (huaccamullu); verde se corne con vinagre y sin ello, y cocida, con came es muy buena y para dias de cuaresma asimismo'-

24' .--yahe dicho que en esta tierra los indios naturales della se sirven y han scwido para su sustcnto dcl maiz y frialcs y altramuccs y papas y camotcs, quc son batatas, y de unas yerbas que ellos ilaman guaca-muilosW.

25' ... Y comian maiz y chuno, papas, quinua, oca y yerbas, que Ilaman yuyos, ..."

26 "-..y le servia y le daban indios, que Los cargaba para cerca indios de Callauaya, para lcjos lucanas, y Ic llamaban Lncapchaquin y hasta dadc carnaos, maiz, papas, aji, sai, lana, dgodon, pescado y camarones chiche conejo hasta darle uacamullo, ocororo, ancaua, llullocha. murcoto, llachac onquena; estas cosas seMan de m%my no daban tanta pesadumbre como ahom.." uacamullo refers to in this case is through its association wïth the other items which are listed after it - "ocororo. amaua. Ilvllock murcoro. and Ilachac onq~ena"~aiI of which refer to types of aquatic plants or waterctress. Trying to iden- the herb known as gtcarcumuIIzi with the intention of associating it with mulh means idehga plant that fits into either the colour associations or ritual associations of the shell. In Historia nady moral de las Indias. José de Acosta (1962 [1590]: 246) describes the use of coca for sa~nficecallïng it an herb. Coca was known by this name at this time by the Spanish but it was not cooked or eaten with Wiegar- Nevertheles- Acosta lists its use in Pem alono with corn- coloured feathers and treasure beads. called mollo. and sea sheils- dl for sacrificiai purposes. Whether coca was combined with mullu to create a guaccnullu is unciear. but appears to be unlikely. Coca is connected to mullu by the necessity of adduig an alkali. in the fonn of lime or quinoa ash to release the aikaloid in the coca leaves. Lime is a form of calcium carbonate. the same material that makes up Spondylus shells. Ground-up Spondvlus calcifer is ohused as lime to aid in coca chewing (Keen 1971) and this association may have been responsible for a conceptual relationship between the two. Definino the tem guacamullu from the information available is difficult. Ethnobotanicd records of modem and post-conquest Pem indicate several possibilities. One herb that may have been referred to as guacamullu is Amaranhus. also known as bledos. and referred to by Bemabé Cobo (Herrera and Yacovleff 1935:84 Fibre IV, Cap. II:337l) and Juan Lopez De Velasco (197 1: 10 11 5741). The green leaves of the bledos. or Arnarantw panimlatus. were eaten in boiled salads and used as a condiment like chili pepper (Herrera and Yacovleff 1935:85). The rad and white seeds (Towle 1961:37)' aiso a source of faod, sound similar to the red and white beads or chaquims of Spondylus shell. Perhaps these seeds represent a fom of pseudo-mdlu due to their physicai resemblance to the chrrquirus. The definition of mulh as "small round seed-iike objects" (Olaf Hoh 1994: personai commimicatÏon) descnk both the chaquircls and the bledos seeds. This identification of bledos as guacamullu may not be correct no maner how plausible. What is more important to the discussion of rnullu than this identification is that the term whïch indudes mullu relates to an herb or fbod with no direct physical link to Spondv2usS If the identification refers to plants hmthe se% such as watercress? seaweed or sea moss (coraiina) that might appear entangled in the Spondylus spines. then an indirect relationship might have existed. However. this is oniy speculation at this the.

Mu& Siuvival in the Coloniai Period and Bcyond

While mullu did not disappear overnight in the Colonial Period. its value as a trade commodïty decreased and its use was outlawed. subject to physical punishment and legai action. During the period when the Spanish attempted to stamp out idolam and native beliefs, regdations were enacted which made it a crime to keep mullu or llacs~among other objects. Those caught breaking this mle received one hundred lashes. had their hair cut off. and faced court proceedings27~~aga1968: 171).

--- 27 "Item. From now on no Indian ofeither sex will keep mullu, paria, or llacsa, oc will make a soncu or tecti, nor will he have an usto. nor keep the corn they cal1 huant- airigua. micsaum or collataara? nor will they keep with potatoes, o- &motes. or yucas? and anyone who breaks this nile will be given one hundred lashes and his hair shoq aud proceedings wdl be suuted qainst him for ha* lapsed into idolatry (Arriaga 1%8: 171). The importance of munir and its ritual connection was not cornpletely eradicated in the post-conquest era. Mullzî appears in several instances as an item found in the tooikits of shamans who were charged as sorcereis and as an item of ritual importance to sacrifice. Around 1700 A.D. in the southem highlands ofColombia in the community of Paccha a shaman known as Andrés Arévalo was charged with king a sorcerer. In his package of magical objects was a collection of pre-Columbian artifacts that included mullu and a "special conch -pet used by mullu couriers" (Salomon 1983: 418). Later. in the mid- eighteenth cenairy town of Andagu& in the Val- of Ayo near Arequipa. the leader of a "faccion de idolatorsWtSebastii Tintap. was similarly charoed with having macal infïuence over his enemies. His llama-skin pack included mullu shell. coca. rnaize, and prehispanic statuettes (Salomon 1987a:lSS). Both instances show not only the mival of the word rnullu but also its connection to matters of ritual sienificance. An even later use of the word and concept of mullu was recorded by Adolph Bandelier who wrote about the Lake Titicaca region of Bolivia (1969 [19 101). He describes an 1895 ceremony performed by the local Aymara on Titicaca Island before being allowed to begin archaeological excavation (1969 [1910]: 97). The shaman who perfonned the ceremony for Bandelier listed the required rihial items as: coca, uira-koa 1- llama-tallow?two fetuses of a llarna and a pi& a piece of the skin of a titi or wildcat grape brandy, wine, and especially mullu. In this case the mullu was in the fomi of a small white animal object. Bandelier concluded that mullu was a fetish of white alabaster representing a bu11 or cow. similar to those found in New Mexico. but also known on the altiplano. In Quechua similar mal1 stone figures were called illus or enqus (Cook 1992:356). The recovery of srnail llama figurines made of Spondylus in archaeological sites may explain how the Aymara came to refer to them as ntuilu. This could be an indication that it was the use and nhial power of the object that was referred to as nrrrllrc. not the material from which it was manufactured. Bandelier( l%9 [Ig 101: 10was unable to determine whether the Indians had other fetishes Iike the one they dednuZZuF but did notice that the animai form fetishes cailed nrullu were sold around the country bv Quichua speakuig shamans known as callahuaya (1 %9 [19 1O]: 103). While muilu was usuaily an animal figure. fetishes in the fonn of men and women were also used (1969 [1910]: lOS.loS> In Peru srna11 stone illas in the shape ofcattle?sheep, and horses are used today as sacrifices to the W~manis~the mountain deities who contra1 the availability of water (Isbell 1978: 151). The pre-Columbian figurines shaped like llamas and alpacas are also identified as illus and are considered to be power generating objects (Allen 198854: Me11 1978: 15 1) that are often sacrificed to the Wamanis to increase the herds of guinacos, llarnas and alpacas (John Topic 1995: personal communication). It is possibIe that the sacrifice of small animal fi-pines. formerly made of Spondylu. to the pis for water is the continuation of the prehispanic sacrifice of mulZu for water. The present day Ayrnara use a similar assemblage of ritual objects. ushg the term resa for certain elements of the ritual complex that include the Lump of llama fat coated with gold and silver paper: stone spirit seats and the cloth on which the anifacts are placed (Sharon 1978:82). In Bolivia the spirit seats are power objects made of alabaster or soapstone again which are provided by the travelling doctors known as qollohuayas (caZlahurryc). Shamans in northem Peru also use a collection of items laid out in an altar- like format known as a mesz for rituai activity (Sharon 1978). ïheir collection of items includes metals? shellsl tobacco, stones and other objects similar to those described in the ritual cemony by Bandelier above. The sheiis listed in the mesa for both the Pem and BolivÏa regions include scailop. peari and snail shelIs as weil as unidentifÎed bivalve shells. In the case of the Peruvian mesa the bivalve shell is called Concha San Juan Bautista and listed as a large oyster shell which symbolizes rebirth that is brought from the ocean (Sharon 1978:168). The tems mullu and Spondvlus are not used but the photographs and drawinp provided by Sharon (1978:84) do resemble Spondyîus. while the reference to rebirth is reminiscent of mullu- The concepts of rituai sacrif5ce may have evolved over time without the presence of mulk or the concept of nzullu may have evolved with a name change for many reasons. In 1942 offerhgs of mullo ntescladizs with the blood of cuy or Uama were still being made to marshes and streams for rain and the feràlization of the earth (Tel10 1967:22). Future research must face the possibility that those concepts represented in the past by mullu may now be referred to by another name, or vice-versa

The difference between Spondvlus and mullu can be summed up as follows: Spondvlus is a shell while mdu is a concept that varies in its context and form of presentation. It is apparent that mullar is more than a term that translates directly into Spondy1usS The possibility that rnullu may refer to Stromb- mother-of-pearl- turquoise artifacts, coloured maize. or even herbs. should indicate that SpondyZus may be just one of a number of items that are involved in the concept of mullu. The gender associations and its position in male/female oppositions, visual representatiom. and use as an offerhg for waar is signiscant in the multi- purpose concept of mullu. As food for the gods and food for the people. the consumption of mullu and guacamullu may have brought more spirinial than physicai nourishment. The coIour associations ofreb yeIiow. turquoise and green with tnullu appear in mythoIogy and riaiaI use. The possibÏiity that wme of these colours represent objects that are known as narZZu~or are nullu-like is dso evident hm the chronicles. 'The recognition of these colour associations and rnullu-iike objects -~ves a strong indication that there is a complex of characteristics or traits that can be ascribed to mulk not just its definition as a Spondylus shell. CHAPTER FOUR

PROCUREMENT, TRANSPORTATION, AND TRADE

Apart hmits rituai value, acquiring mirilu for use in ceremonies created an alternative value for the shell based upon its movement through a transportation and trade network To bruig the shell £kom coastal Ecuador to the Peruvian Coast and Andean highlands meant that mechanisms of procurement and transportation had to be established which could consistently supply the escalating demand. The rituai value ofmullu was based on an ideology that was not af%ectedby distance, scarcity of supply, or tnivel tirne. When sacrificed, the shell's ascribed ideological traits would fiinction regardless of its position in a trade network. Because of this, the ideological value of mullu was independent fkom the value of Spondylus as a trade good. Conversely, the trade value of Spondylus was dependent upon the ideological value of mullu. Whether or not the Spondylus trade involved the assignment of value to the shell in a commercial sense that couid be equated with other objects is unknown. Perhaps there was no commodiîy value for Spondylus shell during the pre-Columbian era. The ideological value of Spondylus may have relegated it to a category of trade quite different fiom that norrnally used for metal or textile goods. To understand the role of the Spondyus trade it is necessary to first look at the methods of Spondylur procurement and transportation, and then to discuss the various trade and Spondylus use relationships that might have been in existence in the preColumbian Andes. While the naturai habitat of SpondvZus rnakes diving the only possible method of acquiring quantities of the sheli. how the diving was accomplished and by whom might not have been known without the recovery of Middle Sich and Chimu artifacts depicting the harvesting operation. Evidence for the harvesting of Spondytlus cornes in the form of metal earspools (Cordy-CoUins 1990: Norton and Marcos 198 1:Mg), textiles (Coniy-Collins 1990:403), and ornaments that are shilar in their diving scene portrayds (Figures 19-24). In most cases a central horizontal object identined as a raft divides the scene into an upper and lower half with two people on the upper half (on board the raft) holding ropes attacheci ta two divers. A simiIar image of Spondvius procurement can be found on the eastem side of the Uhle ciudadela at Chan Chan. where a frieze named Los Buceadores (The Divers) has been uncovered (Pillsbury 1993 :15 1). ïhe Spondylus shell also appears as a design motif on Huaca Esmeraldas east of the city centre (Pillsbury 1993: 1 37). Sheet pld objects fkom the Brüning collection show the divers holding tools for loosening the shells fiom the rocks (Cordy-Collins 1990:399:Lehman 1975 [1924]:2 11. Wooden implements similar to these prying tools were recovered by Max Uhle (Cor*-Collins l99O:399; see Fig. 25) at Site H? Moche Valley, dong with Spondvlus shelis. In some of the representations (Figs. 20 and 25) oblong objects are attached to the divers backs that may have been diving weights. Diving to these depths demands a considerable expenditure of energy based upon the time of descent and tirne needed to mhthe bottom for shells. By looking at a contemporary study of cachidos and fwrados9female divers (daughters of the sa?) on the coast of South Korea and Japan, ( Hong, ReNLie?,and Park l982:2-3)?we can infer the oxygen constmhts and diving Iirnits that are imposed on humans in smiatious similar to that of the buceadores of Ecuador and Pem The massisteci cochido can dive to a depth of 5-6 metres which allows them 15 seconds of time to forage on the bottom. Thejbmdo use divins weights to speed their descent with assistants on the surface to aid in their ascent and to haul up the mater&. These weights minimize the amount of oxygen needed by the diver. leaving them fiee to go to greater depths and spend more tirne on the bottom. The weights allow them to expend les enerw and time reaching the lower depths. Because of this thefunodo can work at depths of 20- 30 metres and average 30 seconds at the bottom gathering shells. Oblong Stones

28 with narrowed centres for tying a rope around and thought to be diving weights were found by Jorge Marcos and Presley Norton (198 1: 148-149) at Isla de la Plata Ecuacior. in the centre of the Spondylus harvesting region. If these are in fact diving weights then it illustrates the organization that was in place for the procurement of Spondvlus. It is interesting that these pictorial representations are found in the remains of cultures fiom areas where Spon&ius did not naturally occur. Diving for Spodjdus couid not have taken place in the Moche or Lambayeque Valley. This therefore implies knowledge of the activity either as it was performed by other groups in distant areas, or by members of their own group who had travelled to those areas to procure their own Spondvlus supplies. A third possibility might be that the motif and use of the symbols were passed dong as part of a rinial complex fkom other areas. This knowledge may also be the result of newcorners to the area already fmiliar with Spondylus diving and harvest. The myth of Lord Naymlap may have been based on one of these newcomers

28 "...Pesos de piedra en forma de torpedos y con garganta en el centro y el ememo para am- una soga aparecen a lo largo de la secuencia desde Machalilla a Manteiio" who brought the Spondjdzis knowleûge to the Lambayeque vaiiey. The legend also includes a host of minor officiais such as Fonga Sigde who scattered

Spondyius powder on the ground before Naymiap as he walked 29 (Cabello Valboa 1951 [1586]:327). 'The 8ss0ciatïon between the lord and Spondyius may have sewed to iegitimate the status of the individual concerneci and mi@ be the foundation for suggestions of Spondvhs use as a status object by the Chhu @avidson 1980). How this knowiedge of SpondvZus divins came to be is unclear. 1s it possible that during certain climatic intervals people of the North Coast had access to Spondvlus for limited time penods? The El Niilo event is the most obvious source of short-term climatic change in this part of the world. Study of climate pattern in the Chimbote area of Peru indicate that fiom 11.000 BP to 5000 BPI the central and northem coasts of Pem were subjected to warm water that today is only found north of Paita (RollinsFRichardson and Sandweiss 19.). The evidence includes archaeological shellfish specimens of species that today are inhabitants of warm water regions Merno*. However. while a warm water event in place as late as 5$ûû years ago might have produced Spondvlus in the

29 "...Dicenlos naturaies de Lanbayeque (y con ellos conforman los demas pueblos a este valle comarcanos) que en tiempos muy antiguos que no saben numerados vin0 de la parte suprema de este Phcon gran flota de Baisas un padre de Cornp&s9 hombre de mucho vaior y calidad llarnado NWapy consigo traia muchas concubinas. mas la muger principal dicese auerse llamado Cetemi mijo en su comptailia muchas gentes que ansi como B capitan y caudillo 10 venian siguiendo, mas 10 que entre elios tenia mas valor eran sus oficiales que fueron quarenm ansi como Pita Zofi que era sus trompeter0 O Taiiedor de unos grandes caracoles, que entre los Yndios estiman en mucho, otra Éiinacola que era el que tenia cuidadado de sus andas y Silla, y otro Rinagintue a cuio cargo estaua la vevida de quel Sefior a manera de Botiller, otro llamado Fonga sigde que tenia cargo de derramar polvo de conchas marinas en la tierra que su Sefior auia de Pisar, ..." archaeological- record Ït is uniikely that it would be responsible for divine scenes appearing in the Middle Horizon - Late Intermediate Period. Ifthe El Nillo was able to afféct the availability of Spo&Ztls in these areas during more ment times then perhaps the images npresent fht-hand knowledge of procurement activity. Anaiysis of the Quelccaya ice cores indicates that an El Nao event took place around 1LOO prior to the diving scene fiezes at Chan Chan being created (PiIlsbury 1993:297).- But the= is a difEerence between climatic changes which lasteci for an extended period of the (6P00 years) and those of the El NiAo which were considerable shorter. The relative brevity of an El N~oevent niles out the creation of Spondvltls beds suitable for harvestino. The presence of Spondylus in Peruvian waters could oniy be the result of Spondylus larvae moved by the El NUao curent. While the mollusk could have survived for several pars it would have been Iimited in sue and unable to reproduce when the cold-water Humboldt current retumed (Sandweiss 1992:152). However, the knowledge of Sponàjlus diving may have difbed down the coast fiom areas as far south as Tiunbez which was stiiI within the shell's native habitat

Transpartation: Ddivering the SLeU

Knowing where the shell originated and how it was hamested is the first step in determinhg how it reached its &al destination. Archaeological evidence provides us with much information conceming trade routes? final destinations, and conte- of use (see Appendix A)* but less about the people and methods responsible for its transportath to those archaeological sites. Through toponyms, cbronicles, and excavation, we can track some of those people respombIe for sheii movemem and distribution. even if we are unsun of the nature of the adexchange. Guaman Poma (1990:1002) uicluded a Mu210 Pongo ou his list of tarnbos situated along the main no&-south road nom Quito to Cuzco in the central Ecuadorian highiands. Whiie the Mullo Pongo idemifi& by Guaman Poma in the Ambato region is not easily located today the appearance of mullu in toponyms is stiU evident. ïhe use of rnullu, nrollo, and mullo. appears in the place names of such present day locations as Mulla Pungu, Mullo Corral, Mulluairo, and others in Ecuador (Institut0 Geopfica Militar dd: 117 1,I 190; see Appendix B). When located on a map, most of these locations are found in the Andean highlands dong a north-south route through Ecuador (see Fig. 26). Place names that refer to Spo~lusdoon (gateways) or Spondv1u.s sites in the highlands indicate that they were placed on routes of transportation. The possibility exists, the% that these locations were related to either the transport, storaoe, or manufacturing process connected to the system of interchange of Spondvlw. In many cases a relationship can be infened between a place and its name to an activity or material with which it had been ~0~ected. ïhis appears to be the case wiîh the word rnullu. According to Albornoz (1989 [15851: l87), in the area of Tomebamba (Cuenca) there is a hi11 named Mollotoro - principal guaco of the Caiiares. Possibly an important site for mulh sacrifice: Tomebamba is Mercomected to rnullu through the Mullu Cancha. a palace with wails decorated with plaster of coloured rnullu beads (Cabello Valboa 1951 [1586]:365), built by the inca hmwhich to rule the northem part of the empire (Mu& 1987 [l 5901: 124). The Millu Cancha has been located in the Pumapungo site on the Cuenca Banco Central property. From this location objects of Spondvltls. including a carved llama figurine (enqa), have been recovered (personal observation). There is a comection between the toponyms around Cuenca and the coast. One modem topon.MuiIupune~ appears in EI Oro province, Ecuador, east of Machala where the road enters the highlands on a route to Cuenca It is possible that these toponyms are iocated on an earïy SpondyIus aade route fkom the coast to the Cern Nerrio-Cahr-Tomebambaregion. During the Inca period, the balsas would car* the shelI hmthe Ecuadorian coast to Turnbes where it would be traasported by road thtough the puerta del mduto Tomebamba in the sierra (Hocquenghem 1993 :7O9). The use of mullu related words in the toponyms of Peru is Iess obvious. Books like Diccionario geogrofico del Peni (Stigiich 1922)' Tuponimos Quechuas del Peru (Espinoza Galana 1973) and Toponimius Quechuas de Cajumurca y sus derredores (Puga Arroyo 1971). do not list toponyms with mull- mo- or mullo- in them. Instead the words with the closest pronunciation to Ecuadorian toponyms include a y instead of a II: moyococha moyobamba. moyopampaI and muyu (Puga Arroyo 1971:63-65). In this case, unlike the use by Hyslop and Rostworowski' the variation is not the result of a recent shift to a new orthography. The lexical problems involved with these orthographie variations have already been discussed. Nevertheless. the lack of mullu toponyms in Peru can not be seen as an indication that places associated with mullu activity did not exist, just that they may have been identified by a difTerent name, It should also be noted that none of these toponps appear afong the coast The issue of maritime trade networks (see Chapter 5) is not aided by the mention of any coastal sites with mullu toponyms. Where are the Puerto Mullus or Millu Pongos of the coastai areas that received the maritime shipments of Spondylus for redistribution? To answer this question it is necessary to look to a lanpuaee- - other than Quechua in use dong the coastal areas that may have had its own tenninolooy for the concept of mulIu. Locating evidence of transpodon can be accomplished through the identification of materid remaias, but the people responsible for that rnovement must be identifid thugh sumivin$ historic infiormation- Apart fkom the aiready mentioned Fonga Sigde (Cabello Valboa 1951 [l S86]:327), the only concrete idiotmation available to us for officers or people who controlled this exchange of rnullu cornes from the Late Horizon period of Inca control. in the writing and illustrations by Felipe Guaman Poma de Ayala (1980) we are introduced to individuals who are associated with mullu in two different ways. Fm there are individuals of hi& rank whose name contains the word mu- and second it appears in the name and duty of a specid messenger connected to the transportation of mullu. In the tirst instance Guaman Poma (1980:60) identifies the Great Lord of the Condesuyos as "capac apo Mullo". Later when speaking of the council of captains that advised "Mango Ynga", he identifies another as "Apo Mollo" (Guaman Poma l98O:3îl). Cabello (1% 1:366) also refers to royal capztanes cailed Mullo Cawna and Mullo Pucaru from Collasuyos. There is no indication of how these officers received these narnes or how they relate to mullu. Possibly the apo Mullo was a lord in charge of the overall mullu activity for a certain sector. However, there are diree possibilities associated with these names: the person holds a position based on a connection to maillu procurement and distribution; the appellation Mullo/Mollo is a mark of distinction or nicknarne used to denote status, ra& or an accomplishment; or it is simply a name that happens to represent an ideologicaily powerful object. nie second reference is ta a person who might have camed the shell from one location to another known as the churu mu120 chasqui, the messenger who brings the shell or perhaps hddsthe corn@ of the shell (Guarnan Poma 1980: 323,757). nie association of churu or snail shell with nullu impiies that the charqui brou@ both Strombus (cham)as well as Spondvlus to the hi-hiands. An illustration by Guaman Poma (Fig 27) shows the messenger biowing a Strombus shell tnimpet and describes the chmuI20 charqui as the son of the lord and brinser of caraco& which can be tcanslated as snail or conch shell (GuamanPoma 1980:323). The mullu chasqueros tcavelled Merthan the other chasqui - no doubt due to the long distances hvolved in obtaining SpondvIus - and were identifieci by a headpiece of white feathers. A similar term, mullu chosqui cmnayoc?referred to the tnca official who was responsible for the trafic in nzul~u(Rostworowski 1975: 335. 337 mcisco Falcon 1571: foi. 226 r.]). The Quechua word canrayoc was used in the Andes to describe a person who specialized in a craft for the benefit of the communitv, the state, or religious group. Rostworowski (1975:337) suggests another term for the Chincha merchants who brought the shell fkom the north as mollo hapic camayoc to differentiate their position fiom that of the pescadores. making them a carrier of shell instead of a messenger of sheli. The role of Spondylus and the responsibiiities of the mullu chusqui camayoc have been describeci in the following way:

The trade for Spondyius was designed to maintain the constant suppiy of one of the principal rituai components of the Andean cognitive structure. The ritual sacrifice of Spondylrs under such cognitive structure ~e~edto maintain the need for more Spondylus and in this way the permanent controlling organization (govemment) of Andean Society controlled the flow of Spondylus by also controllhg the specidist production of prestige products manufactured by Andean craftsmen (to be traded for Spondylus). in this sense the Mollo Chasqui Camayoc were the cornptrollers of money in the Andean World (M~Sisnns: lu). It appean that Marcos is confemng a great deai of responsibiliry, and possibly authoriw on the rnirllu chasqui camcliyoc that is not actuaily evident. There is a difference between acting as messengers and handlers of a valuable object and actually controlling the value or intrinsic power of that object The use of the tem comptroller may be misleading. raising the rnullu chusqui camayoc to a position of status that they did not adlyenjoy. There is more evidence to indicate that they were involved in the transportation of shell than in controlling its value. That there were named officers indicates the importance to the state of this shell but it must be remembered that Spondylus was not the oniy elite or ritual materîal with which the state dealto or desired. Because of the ecological constraints on the access to sumpniary goods?successful traders may have been able to tum the situation to their own advantage deriving political authority for themselves hou& control of Spondylus (Mester l990:268). The suggestion concerns the abilïty of traders to become members of the elite group through acquired status based on political and economic control of sumptuary or ritual goods. Certainly arriving in a new social -oup with a suppl of very desirable mullu would have been usefirl for acquiring status for oneself Burials at La Plata excavateci by Dorsey (190 1) were found with Imperia1 Cuzco artifacts. They may have been the remains of the mullu chasqui camayoc who settled in the area to control trading of Spondylus at the centre of Salangane (Marcos l977/78: 114). However, wbile there is more evidence for an Inca presence on La Plata based on the stylistic simiiarities of ceramics and figurines with those of Lake Titi- Pachacamac?and El Angel (McEwan and Silva 1989) there is no concrete evidence that the Inca presence was in fact represented by mullu chasqui camayoc. Identwing Inca people in an area where Spondylus might be obtained does not necessady mean that the twa were connected. People of the Andes were receiving quantities of Spon4vlus long before the Inca created muIlai chasqui c~oc.Whether these officiais were involved in trade as administrators or merely as porters is unclear. Trade officials would on- have become necesSacy if the value of the shell were chan-dg into a market value. In uiis sense then officials may have been used to admïnïstet the trade in mullu.

Any value that can be ascribed to Spondvius was dependent upon the various ways that it was used by those groups who considered it essential for its ritual or surnptuary characteristics. Ritual use created a demand, the lack of availability of Spondvlus increased its value, and the traders provided the supply. Because Spondvlus was consumed through sacrifice, supplies were constantiy in need of replenishment. Nevertheless, the use of Spondylus as an important ritual item, not avaiiable locally to coastal Peruvians and highland Andeans in oeneral' out of necessity creates another use for the shell. lts status in a trade mechanism can be viewed in two different contem. First, it was used as a symbolic accompaniment to facilitate the trade in more mundane goods, or second? it was used as a commodity in a commercial interchange where its value could be equated with other objects. Pablo Joseph de Amaga (1 %8 [162 1]:45), described mullu as a large sea shell that the Spanish and coastal Indians sold to Indians of the sierra- In the post-conquest era a piece of shell "srnallerthan a fingemail" could be bought for four reales (Arriaga 1%8 [l6î l]:M). Why it held such value was unknown to the Spanish. The only obvious reason according to Arriaga was that the Indians would make beads of Sponh,It(s and place them with their huacas. This explanation itself indicates how littie some of the Spanish understood about the meaning of mullu. On the Ecuadorian coast a treasure bead complex existed as a form of primitive currency- Called chaquira, a form of which was made from red and white Spondvlus. it lasted fiom the Formative period to the Conquest (Galvzh Garcia and Bamuso Perez 1986:63 Marcos 1978:1205 Salomon l986:92). Shell beads from the coast were traded along with cotton fibres and salt to the hi-ehiands for copper. The value of these beads as descriied by Cieza de Leon ( 1 962 [1554]: 151) was such that at times they were traded for large quantities of -gold 30. The beads became a type of currency used by the Spanish for trade with the Indians until they had imported enoush glass beads fiom Europe to use in their place (Holm 1966/67:141). The @ass beads may have replaced the chaquiras of Spondvlus and their monetary value, but not the value of mullu as already discussed. The monetary value of mullu when used as chaquira in the north Andean area differed fiom region to region. A unit of currency in the Amazon of 1577 consisted of twentyfour bone beads which equaled one day's labour or the right to spend one night with a woman (Sdomon l987b:66 [Oberem 197 1: 17 I]). [n the Pasto region a single unit, or brcuaf was worth "one-sixth the value of a cotton cloak, or onefourth of the value of a load of red pepper" (Salomon l987b:66). The fat that Spondylus kept its value over an extended period of time can be explained by its role in sacrifice, accumulation, conversion to jewelry, and

30 "...Traenen sus personas algkn adomamiento de joyas de oro y unas cuentas muy rnenudas, a quien llaman chaquira color& que rescate extremado y rico. Y en omprovincias he visto yo que se tenia por tan pnciada esta chaquira, que se daba harta cantidad de or0 por ella" buid offerings, al1 of which took it out of circulation. This constam need for a consistent flow of new shells helped to maintain a system of trade and manufacture (Marcos l986:îO 1).

The Mulli Trade: Bailding A Netwotk

The existence of trade and exchange -stems in the Precolumbian Andes is not at question. Rather? it is the methods of interchange and transponation as they relate specificaily to mullu that require closer analysis. There is ample evidence of land-based îmie routes? road systems? and llama caravans carrying -eoods eom one area to another in the Andes. indicating that trade could have been conducted completely by land if it had been necessary. There is also evidence that coastal fishennen had the abil* and the means to travel by sea to neighbouring communities for the purpose of establishino contacts that might have included the exchange of goods (see Figs. 28-9). What is at question is the ability of native watercraft to traverse long-distance maritime trade routes such as those suggested for the traff.1~in SpondvIusS Moreover? was the demand for Spondylus the motivating force behind an hypothesized trade network that stretched fkom Mesoamerica e the south of Peru (Marcos 1977178; also see Fig 30)? The possibility of a Mesoamerican co~ectionto the maritime network will be discussed in Chapter 5. Where does the evidence of the intent to travel long distances and the capability to accompiish such voyages corne fkom? In both instances the answer for many researchers is found in the report by Bartolomé Ruiz (Shano 1967: 65- 66 [1525]) detailing his encounter with a native raft and the cargo that it carriecl. The raft with a crew of twenty was intercepted about four leagues south of Atacames near Cubo de la Gaieru on the coast of Ecuador (Guinea 1989343 [Shano 18441; Tello 196722 ). hcluded in the translation that folIows is a description of the seized raft and those trade goods that were used to exchange for Spondvius:

This ship that 1 sa. he took seemed to have a hold of almost 30 tons; it was made wah the fioor timbers and keel of some canes as thick as posts tied with ropes of henequen that are like hemp and the upper thinner canes tied with the said mpes? diis was where the people and merchandise were because the Iower deck was weg it had masts and vars of very fine wood and cotton sails in the same size and style as our bats and very good rïgging of the said henequen that 1 say is like hempo and some stones for anchors in the rnanner of a barber's millstone. They brought many pieces of silver and gold for the adonunent of their persoas for exchange with those with whom they went to trade, among which were crowns and diadems and belts and bracelets and amour for the legs aml breastplates and tweezea and bells and masses of beads and rosederes and mhors adorned with the said silver and cups and other drinking vessels; they brought many cloths of wool and cotton and shirts ... and many clothes, ail of hem very much worked with rich decoration, in colours of cochineal and criinson and blue and yellow and ofaii the other colours and of diverse matmers of work and figures of birds and animais and fish and trees, and they brought some small scales to weigh gold like a roman balance and many other things in some strings of beads they had some small stones of emeralds and chalcedo- and other stones and pieces of crystai; and ail this they brought to made for some seashells of which they made red beads like coral (the reddish colour) and wh& that they had the boat neariy full of them. 3

31 "...este navio que digo que tom0 tenya parecer de cavida de asta treyenta toneles hera hecho por el plan e quilla de unas caiias tan pesas como postes ligadas con sagas de uno que dizen henequen que es como caflamo y los alstos de otnis cafias mas delgadas ligadas con las dichas sogas a do venian sus personas y la merCaduria en heaxuto poque 10 baxo se bagnada traye suys masteles y amenas de muy fiaa madera y velas de algodon del mismo talle de manera que los nuestros navios y muy buena xarçia del dicho enequen que digo que es como CaI18moe unas potales por anclas a manefa de muela de barvero. trayan muchas pieças de plata y de oro para el adomo de sus personas para hazer rescata con aquellas con quyen yban a contratar en que yntervenyan The above reference to red and white beads made fiom seasheils has been cited as a reference to Spon&IusI the thomy oyster? in many articles (Anawalt 1992; Cordy-Collins 1990; Guinea 1989: Hosler 1988; Marcos 1977/78; Mester 1990: Murra 1975; Rostworo~1977). It is hard to find an article or discussion of Andean maritime trade that does not trace the proof of such activity back to this pdcular report Although mullu was not specificaily refend to on this first encounter? the definition of muRu (Gonçalez Holquin (1 952 [1608]); Santo Thomas (1 951 115601) and its use as red and white chaquira does susgest that the sailors or traders were carryïng Spondvlus material in their cargo of goods. The importance of Spondvlus to the Ecuadorian traders can be seen in the value of goods that they were willing to exchange for the shell. Whether they were accurnulating sheli for their own purposes or for nade elsewhere is not clear. That Spondvlus was an object of exchange between Ecuador and Peru can be clearly identified through archaeological remains dated back to the Precerarnic penod at the sites of La Paloma and Aspero (Richardson. McConaughyo Heaps de Pefia and Décima Zamecnik 1990) on the Pemvian Coast. However? it is not until the Initial Period (1 800 BC - 1100 BC) and Earlv Horizon about (. 1100 BC - 100

coronas y dyademas y cintos y puiletes y armaduras como de piemas y petos y tenaçelas y cascaveles y sartes y maços de quentas y rosecleres y espejos ~oarneçidosde la dicha plata y taças y otras vasijas para veber trayan muchas mantas de lana y de aigodon y carnisas a aijdas y alcaceres y alaremes y ouas muchas ropas todo 10 mas de110 muy labrado de labores muy ricas, de colores de matla y carmesy y hanil y hamarillo y de todas otras colores y de diversas maneras de labores e fipas de aves y arnyrnales y pescados y arboledas y trayan unos pesos chiquitos de pesar Oro como hechura de romana y otras muchas cosas en aiginas sartas de quentas venian algunas piedras pequeilas de esmeraldas y caçadonias y otras piedras y pedaços de cristal y anpe todo est0 trayan para rescatar por unas conchas de pescado de que ellos hazen quentas coloradas como corales y blancas que mancasy el navio carsado dellas". (Simano 1%7: 66 [152q). BC)that Spondylus Wns to appeer in the centrai Pewian hi-phlands (Pauisen 1974:601). The nature of the type of exchange taking place is unciear. but the existence of Spont&lus remains may confïrm the presence of trade, or a trade network- The possibiiity of a long-distance network of maritime nade is based on the fact that specinc items of exchange are found in discontiguous areas along the Pacific Coast (Padsen 1977: 153). 'Ihe need for exchange with cuituraily distinct groups of people, in this case people from the Coast of Ecuado- created a trade system (Wilcox 1986:32) that codd not only supply the required shell but also created a vehicle for the exchange of other iess prestigious goods (Marcos 1977/78: 1 15). By examinino the fluctuations of trade over time and the forming and reformino of economic relationships between polities, the study of Spondylus trade can be used to trace the socioeconomic development of the polities involved in the loris-distance trade relationship (Bach 1980:30%8; Norton 1986:133). Conversely, the study of Spondvlus should also be useful in determinhg how diat socioeconomic development affected the trade itself and the people involved in its operation (Schomnan and Urban 198752). Does an increase in socioeconomic development reflect an increased demand for ritual goods such as Spondvlus? The communication provided by the goods, and how their symbolic meanings affected the interactions and cultural exchange of different socioeconomic groups: is as important as understanding the act of trade itself (Schomnan 1989:6 1; Schomnan and Urban 1987:49). Did the trade in Spondylus affect the relationship between trading parnien, creatino a hierarchical relationship based on who acted as supplier and who was the receiver? Spondvlus was valued for different reasons by the traders and the users. The value to the user was based on the belief -stem of their own pup. The value to the trade system may have been due ta the value of the opposite exchange items, or the act of piggy-backing other rcade goods dong with the Spodylus. When considering niullu as a trade item we must remember bat ultimately its value is derived fiom its ideoiogicd sipificance. As social formations in the Central Andes progressed from small independent groups to larger States and final- empires' the rote of Spondylus changed. The mechanisms of trade and/or procurement evolved a~cordingly~increasing the ability to import and expon larger quantities of Spondvlus across larger distances-

Maritime or Land Trade: Articulnthg Systems?

There are only two possible methods of tmnsportation in this area: by land or by sea. The mechanisms of exchange couid vary depending upon the method of transportation and geographic region through which it was travelling. Sponùyhs distribution might have been handled duough a maritime -stem of exchange alone the Coast? where trade specialists were instrumentai in the formation of commercial exchange (Ronworowski 1977a: 182). In the highlands a sponsored exchange -stem based on state controlled redistribution and venicality would have created an alternative system that was not based on commercial value (Marcos lW7/78: 1 19; Rostworowski 1977a:1 82). The understanding of these differences is based not just on archaeolosjcal evidence of Spondvlus. but also on the more complex issues of exchange and goods redistribution that affected the entire Andean area. The presence of these digerent -stems articulating in a trade network could have fonned an exchange system that served an area as extensive as that covered by the archaeologïcal distribution of S'on&ïus- However? whiie the evidence of Spon@vlus distribution cm be used to infer exchange it does not indicate direction of travel. Meritune trade wouid see shell being moved dong the coast and then into the sierrq while land trade down the Andes would branch out into the vaileys travelling towards the coast. Whether the shell was being moved fiom die coast to the siepor sierra to the Coast. is more dificult to determine when deaiing with unworked specimens. Direction of movement is best observeci by evidence hmsites that uidicate the shetl was being worked as it was aansponed down a particular route - such as the cut-off hinge sections found at Ricapiaya near Thnbez in Perm Because of thkt the presence of Spondvlus remains can only play a small part in the detemination of trade networks, If SpondyZus were moved through long-distance trade routes, either overland or by se% then it wouid have passed through several different polities. As polities advance through the stages of political and economic developmen~ correspondine stages of exchange have been identified that might inciude one or more methods such as reciprocity, redistribution, and market exchange. The redistributive economy of archaic States is characterized by the longdistance nade of lightweight goods that include precious or prestige items while trade in the bulkier basic commodities is underdeveloped (Whitecotton and Pailes 1986: 185). Intemal and external vade networks are govemed by different mechanisms depending on the material transported and the distance travelled. Extemal trade in the case of Spondjdus is instigated by its non-availability in areas that require its use. It is possible that the trade of Spondvlus. as a symbolic and valuable item- was responsible for devetoping a method of transportation by which other products were also exchanged and transported (Marcos 1977/78: 1 15). The physical ske of the trade netwock and the te-ON that it covered is important to its study. Of particular importaince is the long temporal duration of a netwark despite changhg socioeconomic relationships. In the case of SpondyJw. there is proof that a long-distance network of exchange existed prior to the time of Spanish conta* beginning in Ecuador and eventuaily reaching as far south as Chile. It should not be assumed that a lonpdistance network of trade was operated by a singie group of traders duaughout the network. It is entirely possible that the network involved a series of short-distance relationships moving objects hmone pupto the next on down the line. Spondvlz~scould have been involved in several dEerent mechanisrns of exchange depending on which stage of the trade network it was travelling through. There is no doubt that a land-based network of trade was used in the Andes. or that coastal cultures had short-distance seafa~gcapabilities. There has been considerable debate among archaeolo~istsabout whether prehistoric coastal peoples of northwest South America had the technology for long- distance sea travel. The evidence of deep sea vessels capable of carrying out maritime made to Peru and possibly even Mesoamerica fkorn Ecuador is archaeologically thin- but historic sources provide some information that indicates sailing rafts were in use along the Ecuadorian Coast and the Gulf of Guayaquil (Edwards 1969:4, 1972860; Zeidler 1978:23). It is known that the Mantefio traders from Ecuador had access to balsa log to build rafts capable of extended coastal and offshore travelling (Edwards l969:8; Norton 1986: 13 1). The rafts were constructed out of seven to nine balsa-wood logs of varying Ien-@hs lashed together with cord to form a pointed prow and a square stem. A mast was stepped in the middle log while a platform or shelter built on top kept the cargo dry (Rowe 1946:XO;also see Fie. 29). The balsas and the one-person caballitos made fkom totora reeds could be used for fishing divins, and coastal uavel, important activities for any maritmie based group pig. 28). A probiem with the use of balsa rafts for long-distance travel was inherent in the bdsa material itself. While the rafts were available for use in codmps they would eventually becorne waterlogged and need to be dried out before being used a-@n. This would certainly affect the travelling the and capacities of long-distance voyages- However, experimentation with pre-contact Peruvian navigation techniques based on historic records indicates that coastal groups had greater navigational knowledg and seafkng capabilities than has been previously accepted (Heyerdahl1957; HeyerdahJ. Sandweiss and Narviez 1995). The use of keel-like guara boards inserted between the balsa logs. raising and lowering them as needed, possibly allowed the balsa crew to travel in any direction at will llkyerdahl. Sandweiss and Narviez l995:27). The information concerning the use of guara navigation did not immediately appear in the contact chronicles, perhaps going unrecognized by chroniclers unfamiliar with the previously unlaiown technique. Fiwe 29: a schematic drawing pubiished in the 1840's by F- E. Paris of the early 15th centuiy balsas in the Guayaquil are* shows the vessel equipped with the guora boards. SaJazarde Villasante ([1568/71] 1992:61 1 reported that the Indians of Puerto Viejo in coastal Ecuador were sendïng wood and planks to Los Reyes Cima) in Peru by sea. Without a technique such as guara board navioation this could have been a the consuming operation as the Humboldt current would have made sea travel south from Ecuador difficult. When travelling south dong the Coast to Peru in the colonial e- the Spanish would stop at Manta in Ecuador. the isiand of Puna. and then at Tiunbez (later at Paita). before continuing on to El Callao near Lima: travellers took on food and fkesh-water at these stops for the lengthy journey !Bor& 1954:35!. Because of the problems associated with currents and winds aiong the Peruvian coas at times the Spanish passeagers wouid disembark at Thbez and travel south by land.

Chincha: The Centre of Tmk?

During the Late Horizon on the southcenaal Coast of P~Ndistribution of foodstuffs and materiai gdsis attributed to the merchants of Chuicha by some sources. The foliowing section wi-II discuss both the historic information which links Chincha with trade and a maritime trade network, and the lack of archaeologicai evidence with which to support S'ondvhs trade activity during the time of Inca nile- The Aviso de el modo que havia en el gobierno de los Indios en tzempo del ingu y c6mo se reparrian lus rierras y triautos.3' transcribed by Maria Rostworowski (1 970)?details much about the maritime relationship of the people of Chincha. The Aviso lias a population of 30,000 taxpayers for the coastal community including 1OQOO fishermen. 12,000 farmers. and 6-0 merchants involved with trade (Rostworowski 1977a:169). The indication nom this census is that in the case of Chincha with 10.0 mbute payets involved with fishing it is unlikely that the la~equantities of fish caught could be used solely by the inhabitants and therefore dned fish was probably the Chincha conmbution to an exchange system (Wallace 199 1b:26 1). The Aviso aiso indicates that the traders

32 An anonymously written manuscript located in the library of the Royal Palace in Madrid. Spain (tomo MCII de "Miscelbea de Ayala" [folio 261-273~1. The date of the Aviso is also unknown ? but by correlating its information with that of other chronicles it appears to have been wrïtten between 1570 and 1573 (Rostworowski 1977: 1O3 [1970]). used copper to buy and sel1 goods.33 This in itseif implies a relationship with the himands where the copper Wokd have orïghated. With 6JM0mercderes to arrange transactions, Chuicha can be viewed as a distribution centre for both coastal and hipland gooads if the account is taken at fwe value. But were the inhabitants of the Chincha Val- involved in Iong-distance maritime distribution and procurement? According to Rostworowski (1970: 144; 1977a: 176) the mercaderes made mps fiom Chincha to Cuzco, Quito, and Puerto Viejo to obtaùi goid* emeraids. and Spodylus for the Inca. However, the Aviso does not actually contain any teference to Spondvlus. instead refeming to chaquiras of gold aad emeralds 34. The sea and its resources are important to any coastal commmïty. and the yunga merchants and fishennen of Chuicha were experts in handling the baisa rafts and caballitos de totwa through which they exploited that resource (Rostworowski 1977a: 168). However: conducting longdistance trade from a One-perSon caballito would have been difficult. Aoain. the Aviso used by Rostworowski (1977b: 137) as evidence for Chincha as a trading centre does not state that merchants were using the rafts and caballitos, only that the pescadores each used one. It is possible that with large- balsa rafts Chincha traders could have made the voyage to Ecuador, but support for that hypothesis must corne from sources other than the Aviso. Rostworowski ( 1977a: 175) also suggests the

- - 33 ..compraban y vendian con cobre" (Rostworowski 1977: 176).

34'~viaen este p valie de Chùicha, seis mil rnercaderes y casa uno de ellos tcnia razonablc caudal, porquc cl quc mcnos bat0 tcnia trataba con quinicntos pesos de oro y muchos de eiios trataban con dos mil y tres mil ducados; y con sus compras y ventas iban desde Chincha al Curco por todo el Collao, y otros iban a Quito y a Puerto Viejo, de donde trai'an mucha Chaquira de Oro y muchas esmeraldas ricas y las vendian a los caciques de Ica que eran mui amigos de ellos...@O stworowski 1977b: 1 38). possibüity that the raft encountered by Bartolome Ruiz was from Chinch and that it was the natives from this craft who gave Ruiz Somation on Chincha and its riches, based on an annotated map35 which referred to Chincha. There is no evidence to indicate tbat this was the case, but if the Chincha people did need SponàyIus for aade then they wouid bave had to obtain it from the wam waters to the no& If different zones or regions of markets and trade had their own basic products of exchange?then the speciai commodity of the northem zone was Spon&us(Rostwomwski 1970:152,1977b: 1 18). The reconstruction of trade pattems by Rostworowski indicates that the role of Chincha in the Spondvhs trade was to import enough shell to exchange for highland materids such as copper. This exchange made the acquisition' processing and exchange of Spondvlus a major economic activity in the Andean area (Hoder 1988:842; Rostworowski 1970: 120). Dried fish, chili, cotton, gourds2 and chaquiras wouid have gone to the sierra in return for metals, charqui (dned meat) and textiles. Pior to Inca nile the involvement of Chincha in this trade activity is seen by Rostworowski (1977a: 176) as a commercial interchange which occurred during the "florescence of the coastal cultures". With the increase of verticality - the practice of establishino ties to altitudinally diverse resource zones by a related group of people - during Inca dezthis commercial interchange by the Chincha merchants diminished. However, the Aviso does indicate that merchants were still a large part of the Chincha population at the time of Spanish contact. Questions concemino the validity of Rostworowski's interpretation about the role of Spondylus have been raised based on the lack of physical evidence in the Chincha region. The evidence of pre-Inca Spondvlus use on the north Coast

35 "...elPuerto y provincia de la ciudad de Chichay "(Rostworowski 1977: 175). of Peru appears far greater than that in the archaeologicai record of Chincha where its presence is exclusiveiy related to the Inca pend (Sandweiss 199210)- If the Chincha mefcbants were involved in Spondylus trade priot to the Inca period then where is the evidence of its use? Anne-Marie Hocquen-iem (1 9931706)suggests that the mail amoums of Spondyius found at Lo DemaS by Daniel Sandweiss (1992) indicate that there was no maritime Spondvlus maflic fkom Puerto Viejo to Chincha36. As a major neutrd Port-of-Trade, Chincha supposedly held economic power. The Inca would not have wanted to endanpr their access to Spondylus (Rosh~orowski1970: 16 1; Wallace 199 I b:î62)- dong with the various other items of tracie. The presence of a fûlly functioning trade network at Chincha aiready suppiying quantities of Spondvius to the highlands may have initially suited the Inca purpose. Since the hiohland pattern of vertical control would not have brought S'ndvlus to the Inca a different approach wodd have been required. at least until the Inca controlled the northern area of supply (Murra 1975 :265). Unlike the hi@land systems of exchange and redistribution operated by the Inca the Chincha area is known as a port of commercial exchange and economic power. Groups in the Andean hi-&lands and the coastal areas of Peru depended on outside social groups to suppiy Spondvlus During the Late Horizon this dependence may have been demonstrated by the Inca's tolerance of Chincha trade activities in order to preserve their own access to Spondylus ~ostwomwski1970: 128). Any maritime trade from the Coast of Ecuador to the south would have inciuded the merchants of Chincha either by trading directly to the port or by meeting the traders at a site fùrther north. The traders of j6 '' La poca cantidad de Spondylus, encontrada en las excavaciones de Lo Demis, podria explicarseI simplemente: por la inexistencia del aafico de mullu por balsas, desde Puerto Viejo hasta Chincha" (Hocquenghem 1993: 706). Chincha would have relieci upon the participation of highland populations in a network of llama caravans that moved the goods obtained through complementarity. The trade of items of localized occurrence Iike Spondylus, Strombus, guano, and arsenical copper would have been based on the cooperation of the highlanders (Shimada 1987:144). But if this were the case, then one must wonder why the Inca had nullu chasqui cmayoc to look after Spondylus procurement and distribution. The archaeological evidence (Dorsey 1901) indicates that members of the Inca hierarchy, possibly mullu chasqui camayoc. were stationed at La Plata island off the coast of Ecuador in the centre of the Spondylus harvestïng region, but it is unclear why these officiais would CO-occurwith speciaiized nades. If the Inca state had its own agents in place at the source of Spona'yius supplies, then Chincha may have only been involved with transportation of the shell and not the administration of its procurement and distribution. The mode1 of long-distance maritime trade is not the only possible rnechanism by which Spondylus might have been moved into coastal or highland sites in Pem. Rather than a maritime network with connections from coastal ports to the highlands, a land-based redistribution network and state administered trade may have been more important economicaliy (Lynch 19895; Murra 1975:267). It is possible that both these models existed at the time of conquest (Sandweiss 1988:99). Whether or not the Inca tolerated trade activisy by the Chincha merchants or dealt with it by controlling existing operations rather than eliminating and replacing them is open to debate. According to Sandweiss ( 1992: 10) it is possible that Chincha's long-distance trade actually increased under Inca rule. While Chimu resisted the Inca takeover, losing power as a result, Chincha may have been rewarded for its peaceful surrender with the Chimu long-distance trading rights (Sandweiss 1992:10). Ifthk were the case then the traders mgr have ken acting in a merchant capacity for their local lords anb in tum, their Inca nilers. The problern lies in the interpretation of the accounts such as the Aviso provided by the Spanish. Were they intepreting the situation as it was. or as it fit within their own experÏences and culturat contexts? The Aviso provides a -eood deal of statistical information based on numbers, and as such contains little cultural bias. It is the interpretaùon by Rostworowski that is at issue.

An alternative view by Susan Ramirer-Horton ( 1982: 132) suggests that "the formerlv independent traders may have been incorporated into the empire as state agents. charged with procuring the goods otherwise unavailable within the imperial -stem." While the use of vertïcality might have supplied the Coast with highland goods and vice-versa there would still have been a need for goods that could only be supplied fiom outside the sphere of the vertical associations. Spondvlus would have been one such item that required specialized agents to arrange for its procurement and transportation. This suggestion may be more appropriate when considering the mullu chasqui camgvoc and al1 the other

camqoc designations listed bv Rostworowski ( 1975:335) for the region. Al1 parties involved may have been working for the Inca state carrying out the activity of redismbution of goods to the appropriate resource zones. The Spanish interpretation of coastal-highland interchange could have misconstrued a system of exchanp or mbute between polities where the "porters of the lords". or emissaries. were rnistaken by the Spanish for merchants (Ramirez-Honon 1982: 129). In this case. the Indians mav have been transporting materials for distribution or storage in another area of the lord's jurisdiction

(Ramirez-Honon 1982: 133j. and not necessarily involved in commercial interchange. Like the problem of understanding mullu as anvthing more than just Spondvhs. the Spanish may have had trouble understanding the nature of the connedon between the people of the coast and the sierra as anything other than a merchantkonsumer relationship based on theu own experieaces. The ethnohistoric information appears to indicate that Chincha was an important trading centre. The archaeological proof of such an important trade operation engaged in the movement of Spondvlus would show evidence of Spondvlus in the Chincha area Unfortunately that is not the case. So farthere is a lack of archaeologicai Spondvlus evidence available fiom Chincha to sugoest that the shell was being accwnulated. worked. traded- or even used in local contexts as would be expected of a major centre involved in its trade.

Excavations at Lo Demis in Chincha by Daniel Sandweiss ( 1992: 102) medUV onlv three Spondvlus framents.- Two "bits of spine" and one finished circular bead (Sandweiss 1992: 102) of Spon&dus are not much to base the existence of a trade network on as they could have been carried there in many different ways. Early excavations by Max Uhie in the Chincha Valley did recover Spondvlus artifacts corn various sites ( Kroeber and Strong 1965). However. Uhle noticed that the Spondylus fmds and fine beads were not nom the Late Chincha and transitional Chincha buriais but instead were characteristic of the Inca burials- Because of this he concluded that any trade in Spon&ius did not reach significant leveis until theka period Kroeber and Strong l96H1 ). It should be noted that Uhle conducted his excavations in the 1920's and, as such. was not referring to the long-distance maritime trade network theory of the past twenty years. When contrasted with the quantities of archaeological evidence on the Peruvian north coast the small amounts of Spondvlus in the Chincha area makes its identification as a Spondvlus trade centre questionable. However. the sarne excavations bv Sandweiss ( 1988: 106) did indicate a large amount of fishing activity had taken place in the Chincha area. and also a marsh where the totora reeds necessary for watercraft could be found. The evidence partiaily substantiates the ethnohistoric information on fishing and the presence of materids for building seagoing vesselst but not the Spon&Zus trade. The presence of emeralds in Chincha connects that area with trading centres in Ecuador such as Puerto Viejo where emeralds fkom Colombia were distributed. While this might reinforce Chinchats position as a stop on a Ming route. it does nothing to reinforce the idea that Spondvlus was being imported by Chincha merchants and then dismbuted to the highlands. Questions about the place of Spondvlus in this network must be answered archaeologicallv and through further examination of ethnohistoric records. The question also needs to be asked. "If there were a maritime mide network of the scope sugested. could it have been driven by a singe object like Spondvlus?" Much of the present Iiterature seerns to indicate that the ritual demand and importance of mullzr were the pnm- reason behind Spondvlus trade networks. with most other ob!ects occupying a second- position of importance.

The MesoamericanConnection

The maritime trade network ftom Ecuador to Chincha is only part of the archaeological puzzle. Contact north of Ecuador to Mesoamerica is also seen as part of a long-distance maritime network. The hypothesized Ecuador- Mesoamerica network is founded on the presence of Andean reiated artifacts in Mesoamerica. Artifacts. ceramic style. technical knowledge of metallurp and bronze. artistic depictions of chimaera rnaize flour. the shaft tomb cornplex textile styles and technique. have al1 been identified in West Mexico. Colombia. and lower Central America (Anawalt 1992:122: Hosler 1988:832.843: Marcos 1977ff 8: 1 17: Meighan l%9: 13: Pauisen 1977:141 and 153). From AD 1000 until the Spanish invasion Peruvian Muence is phdm mngin Mesoamerica. It is during this period that a suggsteed longdistance maritime trade organization was in place connecting Chincha on the South Coast to Ecuador and Mesoamerica in a Pacific Coastal network CHoster l988:832). The rationale for a trade network between Ecuador and Mesoamerica is- according to Jorge Marcos (1977n8: 120). that a dïminishing Spondvlus supply in the Ecuadorian area created the need for extemal trade relationships with parmers who Iived within the geographic distribution of the shell. Increased Spondylus use in Peru and Bolivia under Chavin. HuarLTiahuanaco and the Inca expansion increased the flow of Spondvlus to the Andes. and created a demand for additional supplies (Marcos ( I977/78:120). At Dresent there is l ittle evidence to indicate what quantities of shell were available in the Ecuadorian area or how much exploitation would have caused them to become depleted. A shortage of Spondvius shells might be reflected in the size of shells recovered from archaeological sites. An increase in the recovery of immature specimens may indicate that adult shells had been over-harvested. and as such an analysis of shell size fkom dl sites is neces-. Whether SpondvZus was involved in a trade network ma? only be solved by the trace element anaivsis of Spondvlus artifacts in order to detennine their point of or@&. If S~ondvïuswas involved in trade with Mesoamerica then what were the mesoamericans receiving in exchange? According to Hosler11988:852) smelting techniques and fabrication methods related to the production of arsenical copper are possible exchanse candidates. however. West Mexico had its own supplies of copper and chrysocolla at Ameca. and worked copper appears in shafi tombs around 350-500 AD (Weigmd. Harbottle and Savre 1977:21) long before the appearatlce of a Pacific Coastal vade network to the south. Trading expeditions to Zacatula on the Rio Baisas in West Mexico fkom "somewhere" (West 196 1: 133[Albomoz 152q) south of Mexico are one line of evidence which suggests the possibility of maritime trade. The purpose of these expeditions was to exchange "exquisite thingsn for locd goods. The definition of "exquisite thing" is unclear but shodd at least irnply the trade of finished goods over raw materials. There is no archaeological proof that the somewhere south of Mexico. was in fact Ecuador. Rafts coming from Ecuador through open water would have covered a straight line distance of approximately 3200 kilomenes across open sea (Anawalt 1992: 122: Meighan 1969:12). The possible evidence that this kind of navigation knowled- was in use by native seafarers at the time is restricted to pre-Spanish visits to the Galapagos Islands (Heverdahl and Sk#~lsvold 1990 11 9561). Contact evidence with the islands is based on ceramic styles related to Coastal Tiahuanaco. Chimu. inca and coastal Ecuadorian ongins. and that cotton growing on the islands belongs to a species cultivated on the Pemvian coast (Heyerdahl and Skjialsvold [1%6] 1990:721. The shonest distance between the Galapasos Islands and the mainiand is approximatelv 1.000 kilometm. A certain degree of navigationai knowledge can be assumed bv this evidence of repeated voyages. The ability to carry out these joumeys would rely on the use of the para board techniques previously mentioned. What is not known at this time is the date at which para boards may have corne into use. The knowledge of navigation and the ability to sail into the wind does not mean that sailin8 the open seas was a simple maneuver. Even the Spanish initiallv sailed south dong- the coast from Mexico to Ecuador until a new route following a course shaped like a "fishhook" through open sea was discovered in the 1540's (Borah 1953:3 1). Ethnohistorically known traders from the south of Mexico and ships travellino north off the coast of Ecuador and to the Galapagos Islands are not proof that contact of a trading nature was ever acaialiy made- Travel north to Mesoamerica dong the current would not have been diffi~cultand could be accomplished in four to six weeks - El Cdao (Lima) to West Mexico - during late summer (Borah 1954:30). althou& travellers were at the mercy of the storms that hit the unprotected western coastline of Central America. Unpredictable weather conditions and seasonal ocean cments Kraig and Psuty 1968: 15- 18: McEwan and Dickson I978:365) may have forced traders to stay in Mexico for five or six months at a tirne. thereby explainine the introduction of Andean materials to the area (West l96l:l34b and also the possibility of trading colonies of Ecuadotians living in West Mexico (Anawalt 1992: 1271. A return joumev fiom Huatulco. West Mexico. to Peru took colonial sailors at least seven or eight rnonths during the months of April to October (Borah 1954:301. Contact through land travel or coastal vovages is not an unredistic possibility. however time-consuming it mav have been It is the establishment of a long-tem trade relationship based on Spondvlus that is the primary question here. One of the common traits between Mesoamenca and Ecuador was the use of copper monev-axes that originated in the Andean area and moved northward. The distance between the Andes and Mesoamerica is shortened by the discovery of monev-axes in Nicaragua (Holm I966/67: 142). and the recovery of Stone camelid effigies. a jadeite pendant and a reptilian-like figure from the Las Huacas site in Costa Rica (Fonseca and Richardson 1978 1. Evidence of this kind might indicate that contact took place as a result of short-distance contacts in a down the line fashion rather than a long-distance jump. If there is evidence of Ecuadorian influence in the West Mexican area perhaps it is because people travelled north and settled there due to the diffi~cultiesin returning home. Observation of Ecuadorian style in textiles and ceramics of the Mesoamerican region do aot necessarüy indicate trade as much as thev indicate the presence of Ecuadorian travellen. What the evidence does not indicate, is whether the SpondyZus used in Mesoarnerica is Erorn Ecuador. or whether Mesoamerican Spondjdus appears in the ceneal Andes. The movement of shell and the identification of SpondvIus as an exchange commodity was not unknown in Mesoamerica. Spondvllus. called reorlchipuli in Nahua meaning "divine conch". evoked the planet Venus or "Sefior de la Aurora". and was associated with otiendas- used as grave goods in human intennents. and occurred in caches with and without other material culture (Kolb l987D). Regional shell exchange networks that carried Pinctuda mazathtica and Spondvlus princeps. among other shells. hm the Mexican Coast to the highlands had been in existence since Early Formative Oaxaca (Pires-Ferreira 1982:3 13). A sustained Spondvlus trade relationship is not vet evident in the archaeoiopical record. The need for al temate supplies to meet Andean demands has not been proven. and the presence of specialized mullu traders ftom the Andes working in Mesoamerica has not been mentioned in the historical record. While there is ample evidence availabie to indicate that indirect contact between Mesoamerica and the Andes did take place- direct contact through a long- distance maritime trade network can not be proven at this point. However. the present lack of archaeologica1 proof should not be seen as an indication that maritime mde did not take place. There are many indications that contact did exist. If it did take place it cannot be proven that Spondylus exchange was a part of either form of contact. although the identification of Spondvlus unicoior in Andean sites would ptovide a mong first step in corroborating the theory. There were also enough traders and merchants working in the Mexico-Ecuador comdor who could have facilitateci the transfer of gmds in both directions. If the assumption that Spondvlus was traded fkom Mesoamerica is correct, then more information is required conceming the possibility that Mesoamerïcan proups had a desire to mietheir own supplies of ritually important Spondvlzîs. b

System Operators: Tirde Spcciaüsts

Who were responsible for the operation of this Pacific Coastal trade network? Were independent merchants involved for their own private gain or were the States and polities responsible for the maintenance and operation of the network? The answer to the question is dependent upon the historical period considered. As the ?stems of trade evolved and grew over time. the period best suited to analvsis is the era immediatelv prior to contact with the Spanish . The traders operating at that tirne included the pochteca and ortomeca. professional merchants of the Aztec culture in Mexico (Carrasco 1983:M). the Mantefio traders of the Manabi Coast in Ecuador (Mester 1990:ll). the mindalaes of northem highland Ecuador ( Salomon 1 978:23 7). the Chincha merchants of Peru (Rostworowski 1977:99). and the mullu chqui cumqoc of the Inca state. All of these traders were involved in travel within their own areas, some of them we know were involved with Spondvlus- while others administered trade for their state. If there was a network. then these were the people responsible for its operation. The pochteca of Mexico served the Aztec state bv exponing items made by artisans to ports-of-trade and other locations for exotic goods in a market svstem (Canasco l983:73: Salomon 1978:235). When serving in a domestic capacity the? were pochreca. but were calied oztomeca in distant areas where thev were in contact with other members of die trade network [Camisco 1983:72). In the northem Andes. the control of goods came under the juridiction of the local elite leaders of cornplex chiefdoms. who were not under the conaol of a single panunount leader Mester I990:267). The mindalaes of northem Ecuador were dso a merchant caste who, like the pochteca. were working for political elites as import-export operators responsible onty to their own chief (Salomon 1978:214). When the Spanish arriveci in the nonhern Andes they found gatherings that thev called tianguez or tiangueces. derived fkom the Nahuatl word tlanquzrrl. used to describe a native marketplace and based on their observations of the similarities in trade taking place (Salomon 1978:2391. Also seen at these markets were the red and white chaquira beads made of Spondvius that connect the highiands to the coast. and in turn to the maritime network of made (Salomon 1978:237'). Highland pups also required other raw materials for lu- goods that were used to express rank and relied on access through trade specialists to lowland and maritime tropical zones controlled by foreip pups (Salomon 1978:2331. There is no evidence to indicate that these groups of traders had contact with one another. or that they made a conscious decision among themsehes to fonn a network of trade for mutuai benefit,

There are several points to consider in the study of nade that can be used to indicate patterns of exchange (Adams 1974:241). By approaching these as a checklist of Spont-&lus based activity. it is possible to icientie the ievel to which Spondvlus made had developed. Initial- it is necessary to identi- the source of Spondvlus- which for reasons of a biophysical nature can be Iocated on the Coast of Ecuador or funher nonh. The source of the Spondjdus aade has been placed in the same location by ethnohistoric accounts CHosIer l988:832Mima 1%3:8OS: Simano 1965:6548) and archaeolo~cdendence (Mester 1990: Norton 1986). Second an andysis of the traded materiais for function. association with other artifacts. iconography. and stylistic description should be able to place the item in context Spondvlus is in ail cases used in either a ritual or status function. associated with other shell artifacts and sumptwuy goods. or found in sacrificial or burial contexts. Spondvlus. as a marker of social status and public ritual. is a part of the more complex aspects of trade relationships. Its position in the network was important enough to the Inca to assign an official. the mullu chasuui cumquc- to oversee its procurement and distribution. Its value as a trade item must be measured by the value it retains or obtains at the end of the network. nie trade of Sponms provided a positive correlation between the distance traveled and its value in the network by maintainkg a hi& rinial value (Zeidler 1978128). The purpose of exchanging sumptuary or ritual goods must be considered in relation to the exchange of subsistence and ordinary goods. W-as a need for subsistence goods the reason for exchanp in items like Spondvlus? Subsistence goods would not have tnivelled the long distances that Spondvlus did. but both items may have travelled the same routes at simiiar times. Zeidler 11 978: 131 treats surnptuary and ritual good exchange as a systemic regulator. a method of keeping the trade in subsistence goods moving when either the balance of trade is altered. or the stability of the network is affected. Wright and Zeder (1977234) also view the movement of rÏtual goods as svstemic regdators where the producers of certain soods are not very aware of each other's existence. deaiing instead with made specialists. Another point to consider is the diffince between the socid needs of the of the agents exchange- As Ceroup and those of social groups evolvedr the levels of stratification dso hcreaseà. The practïce of Spondylus procurement became more specialized within a group to the point where those people involved with the operation of speciaiists and distnhtion can be seen as elite groups. ïhe elite groups controlled the economic process that preserved the trade in Spondvlus. Importeci wea1th fkom outside the socid -stem may be wealth that the local elites were fke to amass (Schortman and Urban 198752). but in a system of reciprocity and redistribution Iike that of the Andes- the elites had a duty to the other memben of their society. Spon~swas not just an elite good to be hoarded as a marker of status. It was dso a ritual item which, when sacrïficed. affected the lives of al1 group mernbers. By using the Spondvlus for the people. the elite groups were using it as much for themselves. As Netherly ( 1977266) explained. "on the nonh Coast tt wouid have been the lords who had access to and interest in the cargo of the raft captured by Ruiz. Their responsibility for the spiritual welfare of theù subjects would involve them with the need to obtain adequate amounts of Spon&ïus." Whether the Ecuadorian mindalaes. Manteno traders, and Chincha merchants were oblivious to the possibility of personal gain through the trade and exchange system is not indicated in any historic accounts. While some worked for the state. others may have worked for thernselves. Trade of elite -noods may have brought more benefit to the traders than trade in non-elite g~ods. The question. ";Que organizacion politico-economica mane-iaba la extraccion. transfomacion. transporte y distribucion del Spondylus hacia el sur? " (Murra 1975:258) is being answered through the identification of trade relationships. methods of transportation. and nade officiais. Before mohg on to a discussion of Spondylus disnibution and trade models. it is helpfûl to review what is known about the trade network from archeeologïcai and ethnohistoric sources. First there is archaeologicaI evidence of Spondvlus sheli in rituaI contem at coastai and highland sites in the Andes. Second- these shells were transported fiom the coast of Ecuador through some system. Third. ocean going vessels were available by the time of conquest that couid have made the joumeys required to transport the shell £tom Ecuador to Peru. The encounter of Bartolomé Ruiz with an indigenous watercraft indicates that Ecuadorîan coastal trade was taking place and that their craft were large enough to make deep sea joumeys while carrying a substantial cargo. Fourth. evidence of specialized fishermadtraders and divers indicates the importance of the ~ondvlusde. Fi*. the appearance of cultural traits. characteristics. and other material items in discontinuous areas indicates a relationship between cultural pups based on either maritime trade relationships or land-based diffusion. Much of what is known about the maritime trade network. and the importance of Spondylus to it and socioeconomic development. is based on a combination of archaeological evidence. ethnohistoric accounts. patterns of trade and economic developrnent fiom other cultures. and a raft of twenty native travellers who had the misfortune to corne into contact with Pizarro's conquistadors. The point of origin for these Spondvlus items of trade has been located through ethnohistoric sources and archaeological evidence to areas on the Ecuadorian Coast. As Spon&hs is a non-local item in the Peruvian ana it was necessary to either travel to the source or to enter into an agreement with foreign grouDs who could act as suppliers. The creation of an exchange relationship with a cultwdly distinct goup of peopie initiated a trade system (Wilcox 1986:32') that couid not ody supply the rquired shelL but also created a vehicle for the exchange of othet less prestigious pds(Marcos 1978:1 1%. Shimada(1985:39 1) sungests- that by the Middle Horizon the long-distance trade in Spondvlus hm the Ecuadorian Coast was controlled by the middle Siah eiite with Bath Grande as the hub of the network. The discussion of Spondvlus trade would not be cornpiete without a consideration of the possible Mesoamerican connection to the aade network. The evidence of Spondvlus use in Mesoamerica has been well documented (Moholy-Nagy 1989: Willey 1972). ïhe similarities of its use between Mesoamerica and the Andes are numerous (sec Appendix D for use bv Mavan groupsl. The evidence of contact between the two regions is strong. and the possibility of a longdistance mantirne trade network that inciuded Mesoamerica should not be discounted summarÏly even though at this time there is no archaeologicai proof that Spondvlus was involved. CHAPTER FIVE

SPONDIZUS DISTRIBUTION

The evidence of S,pondylus distribution and its use can be traced through its recovery from the archaeological record. The escalation of its use and contextua1 associations indicates its importance to the various cultures of the Andean area during a period that spanned several millennia. Excavation indicates that Spondylus use expanded both spatially and temporally through the periods we know as the Preceramic to the Late Horizon. From the religious tradition of Chavin, through the Middle Horizon to the oracle at Pachacamac, Spondylus has been linked ritually with Andean cosmology and economic sy stems of trade and exchange (Davidson 198 1:77; Paulsen 1974:605). The ritual position and rnorphological characteristics of Spondylus in the Andean belief system created a ritual value to the end-user that was at some point regenerated as a trade value by those groups responsible for its transportation. The trade value of Spondylus became apparent when exchanging it for other materials. This commercial and ideological value associated with Spondvlus might be considered oppositional, and as such, is sirnilar to the dyadic structures found in the Andean cosmologicaf system. Spondylus and other materials such as Srrombus, mother-of-pearl. turquoise, rock crystal, obsidian, and goods fiom the Oriente have been found in archaeological sites fiom the northem area of Ecuador into Pem (see: Bruhns 1989; Buys and Dominguez 1989; Elera 1993; Grieder. Bueno Mendo- Smith and Malina 1988; Holm 198 1; Shimada 1982). In most of these cases the sites appear in locations that provide ease of access and movement across the Andes. or at least into the Andes. indicating coast-highiand interaction. As an archaeoloaical- artifha Spomfyllus can be defued phvsicallv in two ways: as a whole shell with morphotogicd characteristics that may lead to an identification of species and on& or in a modified fonn showing signs of hurnan alteration. This information may be enough to create patterns of trade and exchange. or to tnick networks of transportation, but mulk the alter-ego of Spondvlus. can best be defined by its context within these archaeological sites and in visual representations. Our understanding of mullu is based on inference. oral history. and interpretation of the context of Spondvlus in the archaeological record,

Archaeological Distributions and the Evidence of Trade

The fmSep in determinhg the relationship between Spondvlus and trade is to locate the temporal a.spatial occurrences of it in the archaeological record. By recognizing its movement in space. a pattern of trade or exchange mav be suggested which cm fonn the basis for a model of socioeconomic activity. The level of Spondvhs use over tirne can be used to indicate the importance of made and the evolution of the svstems of procurement. By combining both sets of information the value of S~ondvZusas a ritual item can be gauged through the amount of time or ene-y expended for its acquisition. The svstem of exchange for Spondyhs began about 3000 BC. (Marcos 1977ff8: 122) and was separated by Paulsen (1 974: 599) into three tirne periods related to the sociocultural status and role of Spondvlus and Strombus: in the first stage- 2800-1 100 BC. shell from the Ecuador Coast was traded to the Ecuadorian sierx in the second stage' 1100-LOO BC, the trading area was expanded south and became firmly entrenched in the culture of the central Andes: duruig the last stage. IO0 BC - AD 1532, the trade area stretched from Quito to Lake Titicaca. As a general indication of Spondvius trade activity. this breakdown into time penods coincides with much of the archaeological information available twenty years ago. However. additional archaeolooical research now ïndicates that the second stage dealing with interaction between the Peruvian coast and northem highiands possibly occurred much earlier than previously thout&. The evidence for these time periods and the amount of use of and distance travelled bv Spondvlus cornes from the archaeologicai excavations of sites ranging from the Preceramic Period to the Late Horizon t see Figs. 3 1.32. and Appendix A). The following sections wiiI deal with the stages and geographic areas of Spondvlus use as set out originally bv Paulsen. by incorporatins more recent data and discussing the nature of the evidence for Spondvlus vade activity and growth.

Ecuador: From Coast to Highland

Spondyhs trade was centred on the Manabi coast of Ecuador. a location where evidence of Spondyhs and Spondvius workshops appears fkom the EarIy Formative Valdivia Phase. through Machalilla Chorrera. and Guangala. to the Manteho Phase CNorton. Lunnis and NaiIing 19831. nie movement of Spondv[us awav fiom the coast begins much earlier. During the Earlv Formative period in Ecuador. hi-land-lowland trade is seen in the large quantities of SpondvZus that appear in the southern highlands at Cerro Nam-o. Identified as a workshop site for producino sheI1 ornaments. Natrio served as a centre for the exchange of manv matenais. indudino Spondylus and crystal CBnihns 1989) to the Amazon basin nathrap 197548). and the montafia cave of Los Tayos (Marcos 1977/78: 1 10.1 14). Occupation at the coastd sites of Valdivia and Machaiilla overlaps with early Cerro Narrio and Los Tayos. (BC 2000 - BC 500). creating an opportun@ for a Spondvlus exchange -stem. On the coast Spondvlus sheli with the red rirn removed is common in Valdivia sites but finished artifact~are rare? suggesting that most cut rim was for export. The most common use of Spondvlus in the Valdivia period was as abraders. poiishers and beads (Meogers. Evans and Estrada l965:37-38). In contrast. large quantities of Spondvlus rim and omaments appear at Cerro Narrio in the highlands of Ecuador indicatino both preparation and completion of the final product (Lathrap. Collier and Chandra 1975:48). The cut red rims may have been traded for coca and other products from the Arnazon basin. The artifacts from Cerro Narrio are interesting not only for the shapes of the finished product. but also because of the choice of material. Cerro Narrio figurines and adomments were carved out of both Spondvlus princeps and cakffer. Unlike the objects manufactured fiom the red rims of Spondvïus princevs. the human-featured figurines at Cerro Narrio are made predominantly from purple calcffer (see Fie. 33). 1 was able to examine the Cerro Narrio Sponfilus collection at the Museo de Jacinto Iijon y Caamafio in Quito (Appendix C)and my comments are based on this sample. Unfonunately. provenience for artifacts in this collection is weak. A major site studv was conducted by Donald Collier and John Murra (1943) but Bennett. Jijon y Caarnario. and Max Uhle al1 re~ortedon the much looted area. The artifacts fkom Cerro Narrio and the surrounding area that make up the Ji-ion v Caamafio collection are tisted genericaliy as artifact~hm Cerro Ndoat the Jacinto Jijon y Caamaiio museum, The Cern Narrio figurine collection represents a unique fonn ofSpondjdus use in the Andean region. No other site. or culture area appears to have manufactureci the fipines medfiom Spondvlus. Today. the figurines from the Cerro Narrio region are Located in many coastal and highland regional museums. but they are al1 attniuted to Cemo Narrio. The manufacturing technique indicates that al1 paris of the shell were utilized in the carving of both large and small objects. The depth of the head section indicates that the hin- portion of the she11 was used along with the outer shel1 for carvine unlike the rnajority of other cultures where the shell's hinge wvas cut off and discarded. In most cases the back of the figure is the unfinished outside section of the shell. Some of the figures are finely featured and polished on both sides. but the thicker figures are rougher and cruder. possibly indicating that the crude figures are an unfinished form. With their chubby cheeks and baby-faced appearance the cmde figures could have been an earlv form of a figure that was worked on and aged over time as its owner aged (John Topic 1994: persona1 communicationl The finely featured and polished figurines may reflect the phvsical maturity of the possessor. the status of the owner. or simply the skill of the carver. The figures range in size. with the smallest being less than a centimetre in width and the largest approximately nine centimetres in length. There is no wastage of material as the diminutive sizes indicate bat the small artifacts could have been made fkom small scraps. It should be noted that the smallest pieces appear to be carved from orange Spondvius princeps shell. There is no indication for why this chan- occurred. but it might relate to a difference in value between the two sheils. or. perhaps the calcifer shell is not as conducive to fine carving. In addition to the fimesat Cerro Narrio. there are complete Spondylus shells without spines, square and round cumtas. chaquirus. pendants. collars. earspools and hi@y polished rim fraements that mi&t have ban a fonn of currency (persona1 observation). Non-sheli artifacts inciude Stone knives. polishers. beads. figurines. bone awls made fiom deer femurs. and large spiral seashells that might be Strombur (Collier and Murra 1943:67-70). The la~e amount of Spondvlus at Cerro Namo and the location of the site on a main east/west trade route indicates that it was a workshop and finishino centre for Spondvlus artifiacts. and that the site was probably involved with dismbution. The Spondvlus figurines at Cerro Narrio were not found in the Late Cornplex. indicating a decrease in contact with the Coast (Collier and Murra 1943:8 1i and possibly the reduction of its importance as a workshop centre. To the east of Cerro Narrio Spondjlus has been found at the cave of Los Tayos. in a valley that provides access to the Amazon lMarcos 1977f78: 110.114:

Porras G. 1978: 11 ). Spondvlus artifacts included trapezoids. perforated discs. circular crowns. cyi indrical beads. rings. a feline mask and a serpent-bird pendant. These last two have similarïties to Mesoamencan mythoIogica1 creatures. and are both also incorporated later into the imaoery at Chavin (Porras G. 1978:37.70). Radiocarbon dates at Los Tavos indicate the Spondvlus was placed in the cave at 2971) B.P. 50: or approximately BC 1020 (Porras Cr 1978 -63) which predates Chavin's Early Horizon New Temple. The most important and impressive artifact fkom Los Tayos must be the feline mask which is now housed in the Centro de Investigaciones Arqueologicas of the Universidad Catolica in Quito. Whether it is a mask or a erget is hard to tell. It does have what appears to be a mouth slot and two holes for eves. but there is a third hole of the same site off to one side that reduces the impression that it is a mask (,personal observation). What is clear is that it was carved from a very large piece of orange Spon&Ilus pdncevs that had to originate at the coast There is another mask presentIy located in the site museum at Real Alto (1 994: personal observation) of a similar sbape with eye holes but no mouth. that is not fuushed or polished like the Los Tayos mask but which might indicate that mask- makine originated on the coast with the shell- Both masks are approxïrnately a0 cm in length. indicating that they were fashioned from two very large examples of the princeps varkty. At the coastal mainland village of Salango the remains of specialized production activities reiated to Spondvlus working were uncovered at site OP-JP- SI-140 on the punds of the Fabrica de Salango (Norton 1986: 140). Salango itself and the Ida Salango were identified by Ruiz (Simano 1967: 66 [l SX]) as an important town for sea trade. The archaeological remains inciude ceramics. obsidian biades. and chisels made from the white centres of Spondvlus calc[fir dated to the Guangala phase at 100 BC to AD 800. Large quantities of Spondvlus nuclei with the red Ni removed date to the Manteilo phase (Stahl and Norton 1987:384). Also uncovered at Salango were quantities of unworked gastropods and bivalves: 29 of the species present could have been used for food. Copper from the highlands found at the site might represent the trade counterpart to Spondvlus. A ma_ior feature of the Salango excavation are ienses of ash and carbon found in association with fiapents of shell and lime often contained by an olla or DO^ (Allan n-d.: 141. These remains may result fiom lime kilns used to reduce the cut-off white cal~iferrims to lime used for coca chewing (Allan n.d.: 14: Mester 1990:27). This removal of rhsappears dong the coast south to the Santa Elena Peninsula from 100 B.C. on. where al1 the coloured margins of the Spon@ius were cut awav and no immature Spondvlus specimens have been recovered (Paulsen 19745971. Evidence of Spondvlus use at this site is undeniable- but Spon@vIus processing was ody a small part of the activity uncovered at Saïango- The multiple levels of occupation and variety of data recovered fkom the site indicate more cornprehensive activity was taking place. Any suggestion that it was pn'marily a Spondvlus workshop requires closer examination (Richard Lunnis. 1994: personal communication). Excavations on the Isla de la Plata 40 kilomemes northwest of Salaneo- indicate that Spondvlus princeps may have been imported and stored on the island for shipment down the aade network to the south Morton 1986:141). Storage is sugpsted due to the modem-day abundanceof calc[feroverprinceps in the La Plata area but the lack of Spandvlusprinceps in modem waters does not necessarily indicate that the same species was not more abundant in the pre- Columbian era- The excavation of site OM-PL-IL44 on the Isla de la Plata uncovered more than 600 valves of Spondvlus princeps that had been cleaned and placed face down in a 4 by 5 metre sealed area Morton 1986: 1411. Norton's suggestion that the stored Sponr;-lus was destined for made to the south is unsubstantiated. Spondvlusprinceps may have been stored for trade purposes or placed in a ritual context. Rituai activity on the island is indicated by offerin- containino cut Stones. beads and pendants of ruquoise. beads of Spondvlus. ceramic and Stone fieurines with the heads removed. These offerinos continue through Bahia Tolita, Jama-Coaque and Guayaquil phases of occupation (Marcos and Norton 198 1: 147: see Fie. 34). The importance of Spont@lus to the La Plata site is obvious. but it is not proof of the xvondvlus role in a trade network. On the other hand. the presence of Spondvlus. corai. and emeralds in Middle Sich elite tombs during the Middle Horizon, and Sican cerarnics on the Ida La Plata does indicate that interaction was taking place and that the Spodylus possibly was moved through a marithe network (Hosler- Lechtman and Hoh 1990:7n. At Los Frailes- nonh of Salango. the Spondvlus artifacts recovered inciuded rectangdar piques. a Sponc@ltlscalcftèr plumb bob- and cuentas made of the red rùn. The care with which the red çuentas were manufactured and the lowerquality of the white cuentus is taken as an indication that the red cuentas were meant for the longdistance trade network while the white cuentas were used local1 y (Mester 1990: 1 77). Funher north at Atacames excavations of tolus by Galviin Garcia and

Barriuso Pérez ( l986:6 1-63) have turned up a large number of red and orange coloured Spondvïus shell chaqiriras. dated to the Inteqation Period (AD 50- 1500). It is in this area of the Esmeraldas coast that Ciscala was located. which was described in 1569 bv natives of Manta to Captain Andrés Contero as a place where there was peace and safety for al1 to corne. and buy and tradej7 (Carranza 1965:89). As a trade centre. Ciscala might have required the use of primitive currencies such as the chaquiras already mentioned. A route fkom the Quito area to the coast can be found through evidence of material other than Spondvlus. but with a Spondyhs connection. Obsidian. like Spondvlus. can be connected to a specific area of origin. in this case volcanic regions. In Ecuador. obsidian artifacts have been found in the Milagro-Quevedo culture area whose people facilitated the exchang of products fiom the coast to the highlands during late Prehistork times (Holm 198 1:3 1). A commercial route

37'' Dicen hay un pueblo grande que se llarna Ciscala que tiene paz con todas las demis provincias, v aquel pueblo es seguro a todos v aili se hacen ferias O mercado y los tacamas arien Oro y esmeraldas a vender y. los campaces y pidres llevan sal y pescado y. los beliquiamas llevan ropa y aigodon v hacen alli sus mercados" (Carranza 1%5:89). from Quevedo to Latacunga may have bmught the obsidian south to the coast in exchange for Spondj?lz~ One source of obsidian in northet-n Ecuador is in the Valley of Mullumica about 30 km east of Quito in the Guamani hi-&lands (Athens l992:2lO: Salazar 1985'). The obsidian ftom this volca~coutcrop is distirquished by its reddish brown colour (Salazar 198%139). By using X-ray fluorescence and neutron- activation analysis. obsidian artifacts fiom El Inga Chobshi Cave (Cuenca region- and Site OGSE-46 on the Santa Elena Peninsula have dl been identified as ori~inatingfrom sources within 15 kilometres of the Mullumica area (Bure. Asaro. Michel. Stross and Salazar 1994:228). investigations in the area have not produced anv Spondvius matenal. leaving us to speculate whether or not the name of the Valley derives fiom the Quichua 38 mullu or if it is related to mullu because of its crystal-like appearance. The obsidian may have been traded for Spondvlus. or at least associated with it because of the red colour of this type of obsidian and its importance as a trade object. If the name is used because of the tramlucent shimmenng quaiities of the obsidian itself. then the word muliu does appear to represent a concept other than simply that of Spondvlus. In this sense the brightness of obsidian would have more in common with mother-of-~earl.not the duIl sheen of S~onddus. It is obvious that Spondvius was transported from the coast to highland centres throughout Ecuador. Archaeological distributions indicate that other materials were moving in the same fashion. but exactly which route was used is still unknown. Spondvlus from the coast near Atacames and Esmeraldas could to ~oingnear Ouevedo. That have been transported east Quito without ever Ci

38 In Ecuador the word Quechua is commonly pronounced and spelled as Quichua. does not mean that Spondylus destineci for Quito did not go through Quevedo, just that it requires more evaiuation. As the last perdof tirne pnor to the Spanish invasion, the Late Horizon is the era that we know the moa about - as seen through the eyes of the Spanish. In the northen area of the Andes the presence of the mindaiaes, trading over long distances, represented a different politicai economy than that of the southem and cenaal Andes (Bray l992:219). The ntindalaes operated in Ecuador and the Colombian Andes trading chaquiras nom the Coast into the highlands as monetary wealth objects (Salomon 1978:236-7). Because of this association with rnonetary weal* or primitive currency, Salomon (1978:Xl) suggests that the economy of the northern area has more in common with Mesoamerica than it does with that of the Inca. An analysis of ceramic materials from northern Ecuador (Bray 1992228230) indicates that the Inca were transforming regional înterdependency into local dependency on the Inca state while at the sarne time attempting to curb long- distance trade activities. Inca control in the northemmost area of the empire was not as firmly entrenched at the theof contact as it was in the south, resulting in partial tolerance by the Inca for the existïng trade system (Salomon 1987b). Recent investigations into the procurement and exchange of obsidian in Ecuador (Burger, Asaro, Michel, Stross and Salazar 1994:228-255) during this period indicate that the mindala operated exchange network might not have linked the country in the manner suggested above. Instead, it appears possible that the northem and southem Ecuadorian groups were not closely linked by interregional exchange, but were separated by sociopolitical and economic pressures (Burge- Asaro. Michel, Stross and Salazar 1994250). This does not mean îhat exchange was not taking place, or that Spondylus was not part of a coast-highland aade networkI just that the politicai economy of Ecuador requires Merinvestigation.

Peru. Over Time And Space

Archaeological Spondvius occurs at sites located throughout the coast and highlands of the Andes. dating to various time periods. The distributions and associations cm best be understood by examinino its appearance through the chronological sequence and identehg inter-site relationships and context of use. In Pem the fim evidence of S'pondvius use possibly predates the second stage period ( 1 100- 100 BC) originally defined by Paulsen ( 1974599). Unfortunately there are no major finds of worked Spondvlus artifacts. onlv fragments with little published documentation. According to Robert Feldman 11993:731. archaeological excavations of sites from the Preceramic Period (3000- 1800 BC) have recovered S'ondyhs at Aspero and La Paloma on the central coast. The quantity and tye of Spondyius artifacts fomd at these sites is not specifically stated. The finds are referenced to unpublished dissertations by both Feldman and Je* Quiiter. Feldman (1982:8 1) refers to a hi@ status burial at Aspero of a newbom infant found on the floor of Huaca de los Sacrificios, that included over 500 beads of shell and other materiais. but does not identi- the shell. It is only later when discussing longdistance trade and the presence of artifacts made corn exotic matgrials that he mentions Spondvlus shell and admits that " although only a single small £kgnent of the red shell was found in the Aspero midden. it came from a reasonably secure context" (Feldman 1982181 ). Feldman does not Say that the Spodylus shell was inciuded among the shell beads found in the infant burial, Richardson. McConaughy. Heaps de Pefia and Décima Zamecnik (1990:437) cite Quiltefs unpubIished PhD. dissertation when referrhg to "a few fragments" of Spondvius from La Paloma and Feldman's for Aspero. but are not more specific. Quilter (1989:24) attributes the discovery of Spondvlus at La Paloma to Frederic EngeI who found the shell in 1973. although its "exact provenience within the site is unknown" (Quilter I989:29). The published information on Spondvlus at these sites is weak. and because of this it is difficult to assess its mie implications. At ben. it appears that there are only minute quantities on these sites. At Huaca Prieta a late Preceramic momd in the Chicarna Valley CBC 3 100 - BC 1300) excavateci by Junius Bird (Bird and Hvslop 1985). there is no mention of Spondvlus shell among the list of mollusks which were identified to the genus and species levels. Excavations by Frederic Engel (1963:80) at Asia another Preceramic site in the Ornas Valley. produced no evidence of Spondvlus at that site either. This does not implv that Spondvlus was not present at Aspero. La Paloma or El Paraiso. just that its appearance was sporadic at best during the Preceramic Period. with no evidence of widespread orgmized distribution. Better documentation appears for the Late PrecerarnicAnitial Period sites of Los Gavilanes in the Huarmev Valley (Bonavia 1982: 143). and in the highlands at La Gal@a(BC 2300- BC 2200)(Burger 199253: Grieder 1988:89-94: also see Fi-. 35 and 36). The presence at El Paraiso of coloured feathers (Quilter 19851 fkom the eastem slopes of the Andes. along with the small arnounts of Spondvlus- indicates that some level of exchange was alreadv taking place between the coast the Andes. and bevond. Lathrap 11973: 177) argues for a trade association between the Upper Amazon and the coasts of both Ecuador and Peru for this theperiod (2000 BC - 1500 BC) based on cenimic styles and the use of chonta wood fiom the tropical forests for carvllig doils recovered on the coast. The trade situation parailels that of Late Formative Ecuador where a laterai exchange existed dong a coast-hi@Iand-Mentecorridor (coast-Cern NeLosTayos). The fim solid evïdence of Pemvian Spondjdus use still appears in the Initial Period CBC 1800-BC 600). The north coast of Pem and its related highiands were delineated fkom Ecuador by an ever-changïng Grontier that separated the south and central Andes from that of the northem area @Iocquen@em 199 1:3 15). Nevertheless. it was permeable enough to dow for the presence of marine shells in both coastal and highland ceremonial centres and wavstations iocated along inianci exchange routes (Hocquen@em 199 1:3 1 5). The movement of Spondvlus south from Ecuador across this fiontier is seen in the Initiai Period at Cerro Naaaique near Chulucanas in the upper Piura Valley, in what rnight have heen a -eatewav for long-distance exchange across the Andes (Burger 1992: 10 1). During the Initial Period SpondyZus is found at several coastal sites. At Puémape in the Cupisnique Valley. Spondvlus beads are found in buriais with turquoise and mother-of-pari (Elera 1993:246). At Punkuri in the Nepeiia Valley (Proulx 198538) and at Garagay in the Rimac Valiev it appears in the form of carved pendants. beads. and whole shells found in buriai contexts and in caches possibly used to sanc- ceremonial buiidings (Buqer 1992: 63.89). The bun*aI at Punkuriincluded a pair of Spondvlus shells in association with a decapitated female while a Spondylu shell bead from Garagay appeared in a votive offe~g in die floor of -id B. In the Earlv Horizon. evidence of layers of ground Spondvlus appear at the Temple of Mono de Eten in the Lambayeque Valley (Elera l993:249). The use of -aound St~ondvZusin rituai contexts occurs in manv Andean ceremonies. but its sumptum use as a status marker dl appear later in the bun-ai platfamis at Chan Chan* Both Spondvlus and Strombrcs have been fond at Chavui de Huhtar. Ln both the Old Temple and the later Janabarriu Platform SpondvZus shells appear in wdl and floor burials and as debris @urger 1992: 138-169.1 7 1). Ceramic vessels based on the stimp spouts fkom Late Tutîshcainvo occupation in the montufia indicate the poaibüïty of trade with Machalilla and the Santa Elena Peninmla in Ecuador- (Burger 1992: 130: Paulsen 1977: 147-8: Lathrap 1973 :1 77). A sirnilar relationship of trade or the exchange of ideological beiiefs with the Paracas region during the Earlv Horizon can be found Ïn the Carhua textiles showing both a C havin related iconography and a Spondvlus motif (Wallace 199 1a 1. Excavations of the cemeteries at Cerro Colorado in the Paracas region have uncovered Spondylas shell beads. necklaces. and bracelets in contexts that indicate Spon&hs was used as a -bol of elite status (Carrion Cachot 194959: Paul 1WO:3 9-42). The obvious ideological importance of Sr>on&lus at Chavin may have been responsible for an increase in the fiequency of its appearance into and during the Eady Intemediate Period. The recovery of Spondvlus shell in the Nazca Valley during the late Eariy Intermediate Period (BC 200-600 AD) and eady Middle Horizon tNazca 8 appros. AD 750-8501 iIiustrates the extent to which trade in &uon&ïus was expandino. At the base of Unit 19, in the Room of the Posts at Cahuachi. a vacant ceremonial centre in the Nazca Vallev on the south Coast of Peru ten complete unworked Spon~vlusshells were found interred in the sand filling a wall niche with two others found in a circular depression nearby (Rodriguez de Sandweiss 1993:294: Silvemian l988:4l7: 1993: 178). While Chavin mav have been responsible for an increase in ideological importance of Spondvlus and visual representations. the use of Spondvlus at Nazca as dedicatory offerings in constniction is more closeiy ailied with the fuidinos of Moche sites on the North Coast nie simîianaritiesof Spondvlus use do not remah consistent in al1 contexts over space and the. WhiIe Moche in the Eariy Intennediate Period was using SpondndvIusand airquoise inlays. excavations by Grieder ( 1978) at the hi-land Recuay site of Pashash in the CalIejon de Huaylas do not indicate the presence of any Spondvius materiai in the architecture or among the buriais ïhis discrepancy is unexplained. It is during the Middle Horizon that the importation ofSpon&Zusshell ont0 the North Coast of Peru increases noticeably (Cordy-Collins 1990:408). Spondvius begins to appear in the burials of Pachacamac in the Lurin Valley (Uhle 1991 [1903!:371 The oracle at Pachacamac was founded in the Early Intennediate Period and lasted until the Spanish conquest. Whoie shells are found in burials in front of the Temple dong with omaments of Spondvius hm the female buriais in the Cemetery of Sacrificed Women. The female burials inciude neckIaces. bracelets. chaquiras. pendants- and zoomorphic foms inlaid with mother-of-pearl. some of which are iniaid with stone identified as lapis luzuli or sodalite (mie 1991 [1903]:95). Further south on the central Coast in the Ica Vallev. Spondvius is found in the tomb at Pinilla (Paulsen 1968.1. Unworked f-ments and drilled Spondvlus objects were buried with artifacts of copper and -eoid. At the Moche site of Pampa Grande in the Lambayeque Valley radiocarbon dates of AD 650 are given for a bumt cane roof fkom a Spondvius workshop adjacent to Huaca II (Shimada 1WO:3 721. Also at Pampa Grande a smali scale copper working area was also identified bv Shimada ( 1987: 137). who concludes that the association between copper and Spont?dus indicates that they were nade objects of comparable value. ïhe combination of copper and Spondvlus might also be the resuk or the reason for. the reiationship of IIacsa and mullu found in the myths of Huarochiri. Other investigations at Pampa Grande (Jonathan Haas 1985: 397,4O 1,404) reveded whole Spondylus shells, pendants of Spondylus, and necklaces made of trapezoidal pendants, some found with the bones of a child and an immature llama skeleton, The association between Spondyh, chiidren, and immature llama skeletons is seen again in the Middle Horizon and Late Intermediate Pend at Cerro Blanco, Huaca el Dragon, Huaca de la Luna, Tacaynamo, Calvario de los Incas (Bourget 1995:2) and Huanchaco (Donnan and Foote 1978) in the Moche Valley. A cornparison of collections from Cerro Blanco, Huaca el Dragon, Huaca de la Luna, Tacaynarno, Calvario de los Incas indicates that "al1 of the five temples contain great quantities of Spondylus and Conus shell, both in complete or in rectangular form" (Bourget 19953). The largest number of shells were found at Calvario de los Incas, along with evidence of the incineration of shells. The incineration of shells along with textiles is considered by Bourget (1995:3) to be an indication of ritual activity, although bumïng the shell could result in a lime source used for coca chewing. At Cern Blanco and Huaca el Dragon, where the Spondylus appeared as "rectangular pieces, complete shells, or in their destroyed state", there were no tools or evidence of workshops (Bourget 1995:3). Excavations by Navarro htander ( l986:27-29,32,40) at Tacaynamo produced zoomorphic and geomemc forms, beads, and inlays, of Spondylus. The presence of Snombus and mother-of-pearl at Tacaynamo (Navano Santander 1986:27- 28,32) does not alter the fact that whiie the quantity of Spondylus artifacts increases over time with the examination of Moche Valley sites, so to does its association with Conus fergusoni. Like Spondylus, Conus fergusoni is on1 y found fkom Baja, California. south to the Santa Elena peninsula of Ecuador. While it cari be found intertidally it more commonly occurs in water up to 165 metres in depth (Keen 197 1:667). A tentative visuai identification of whole Spondyltls at Cerro Blanco and Huaca El Dragon indicates thaî four varietïes ofSpo~lt(swere present (Bourget 1995: personai commu~catian).The importance of this identification cannot be overlooked, for two reasons. Fkf it indicates the possible presence of Spondylus princeps unicolor, believed to be restncted to the Coast of West Mexico, and an association with long-distance trade. Second, whether the identification is correct or not, it is an indication that researchers are now aîîempting b identify to the variety level whenever possible. At the highland site of Marcahuamachuco (AD 400 - 1000) in the Condebarnba Valley, 9.6 kg of Spondylus shell were recovered fiom an offerïng (dated to about AD 600-650 4.Topic 1995: personal communication) fond in a pit near the Castillo (T. Topic 1989:3). A combination of complete half-sheIls and cut rectangular pieces with holes drilled in the centre top and bonom were recovered along with a quantity of cut and carved turquoise Stone, some shaped to look like Spondylus shells (T.Topic 1989:3). At another location on the sarne site more Spondylas material was uncovered simiiar to the first cache but also including tubular beads of Spondylus (T. Topic l989:7). Further south near Cuzco, two collections of 40 turquoise figurines each were found at the Middle Horizon site of Pikillacta in context with Spondylus princeps, Strombus shell and a bronze bar (Cook l992:344). In the Ayacucho Basin, 15 kl northwest of Huari, the site of Azangaro displays a similar use of turquoise and Spondylus in effigies, geometric forms, beads and fragments (Anders l986:2 11 ). The relationship between mullu and bluish-green stones (turquoise) has already been presented in Chapter Two, but the association of Spondylus with turquoise and bluish- green stones appears almost exclusively in Huari related sites. It is possible that the origin, or at least development, of these relationships can be seen in the archaeological record of Marcahuamachuco, Pikillacta, and Mgaro. Like the Moche Valley temples that may have been the foremers of the fiunerary platforms at Chan Chan (Bourget C995:9, these Spondylus and turquoise artifam may have been the inspiration for the related visuai representations and associations of the Chimu, Bridging both the Middle Hoenand the Late Intermediate Period in the Lambayeque Valley, Middle Sich (900-1 IO0 AD) art and ceramics fiom Bath Grande depict divers collecting Spodylus and the Si& Lord holding Spondyius shells in both hands (Cordy-Collins 1990:397; Shimada 1990:326). Tmis decorated with representations of the Siah Lord and animais include turquoise and Spondylus inlays as well as metal representations of Spondylus shells (Carcedo Muro and Shimada 1985:69-70). Spondylus found in the Huaca Rodillona at Batan Grande was placed as fil1 in column boxes, altemating with bundles of 1-shaped arsenical copper foil, or naipes, in association with a human burial (Shimada 1990:341). An estimated 400 whole Spondylus shells were found at Batan Grande buried with 200 human sacrifices and 2,000 bundles of arsenical copper foi1 (Shimada 1990:367). Similarly, in Ecuador copper axe-money, consisting of a copper blank similar to naipes, symbolized value and was also found in tombs (Salomon 1986:93). This is a Merindication of the expanding relationship that existed between Ecuador, the source of Spondylus, and Peruvian cultures. The presence of Spondylus, coral, and emeralds nom Colombia in Middle Sican elite tornbs, and Sich ceramics on the Isla La Plata off the Coast of Ecuador, is seen as an indication of marithe activity that was already in place by the Middle Horizon (Hosler, Lechtman and Holm 1990:77). The extent of actual maritime movement may have been quite small. The shortest distance fkom La Plata to land is only 23 kilometres which means that goods could have been transporteci by land the rest of the way. During the Late Intetmediate PerÏod, Sponâjvltls is found at Pacatnamu in the Jequetepeque Valley. Buriais hcIude Spondylus vaives tied to the hands of a young female (Verano and Cordy-Collins 1986:87) with bumt shell and beads (Bruce 1986:98). Probably the most important occurrence of Spndylus was found in the royd buid plaeorms at Chan Chan where it was used whole, ground or cnished into powder (Conrad 1982:99,104; Dandson 1981:77). Excavations at Chan Chan in the Moche Valley include Spondvlus in subfloor burials in the audiencias (Keatinge 1982:203), a stone-lined bin of ground Spondylus in the forecourt ofa pplatfiorm (Conrad 1982:96), and six complete Spondylus shells in an olla under a kitchen m the SiAR area (J. Topic 1977:W). Further evidence of Chimu use of Spondylus has been recovered fkom sites in the Viru and Chicarna valleys (Davidson 1980:7). In the Moche Valley Spondylus is found as fragments, dust (powdered) and as valves. The Chicama Valley finds at Chiquitoy Viejo appear only as fragments and valves, while the Vini Valley sites (V-124, V-304) only indicate the use of valves (Davidson l980:79). Although the use of different recovery techniques by researchers may be responsible for the different results. The use of Spondylus in a powdered form may have been overlooked or missed during collection (Daniel Sandweiss 1995: personal communication). It is possible that the difference in fonn of Spondylus use at various Chimu sites follows a hierarchy of Chimu settlements by rank. Mackey (1987: 1%) ranks the Chimu sites into four levels according to the presence of audiencias, patios, storerooms, burial platfonns and number of compounds. The use of Spondylus at these sites can be ranked in a somewhat similar fashion Site V-124 with only valves evident falls into the category of an administrative centre ranked on the third level of Mackey's list. Chan Chan as the primary centre is the only site where powdered Spondylus appears. Not only was the use of powdered Spondylus resefved for the ruiing group at Chan Chan, but Ït also indicates that gound or powdered Spondyllus was associateci with eiite status as a sumptuary good. Looking at the use of Spondyius throughout these time perïods shows an increase in the use of whole sheIls and variations such as crushed or ground shelis, in dedicatory offerings fmm the Early Intermediate Period to the Late Intermediate Period. A decrease is noted in the use of Spondyh as pendants and necklaces over the same time period. Spondylus is used as a niw material in Chimu mark, as inlay materid. But it is also fkequently represented as an iconographic object manufactured fiom other media, especidly ceramics and metalwork. These representations usually depict the entire shell. This might indicate that a greater importance was attributed to the shell itself than in any status associations that could be derived fiom wearing small pieces of Spondylus in the fonn of pendants and necklaces publicly. Davidson (198059) argues that Spondylus did in fact imply rank in Chimu society based on the restriction of the shell to elite groups:

"By manipulating the contexts in which this shell (Spondylus) is represented, the Chimu artisan cornmunicates the elite control over the distribution of status cornmodities, The close association between this shell and the highest ranking individuals in Chimu society is visuaily portrayed by the iconographic device of limiting the manner in which this shell is shown aith human figures. For example, aldiough it is shown with various groups of the Chimu elite, this shell is only held in the hands of the most elaborately costumed figures. When portrayed with lesser ranking individuals, the shell is represented as the spouted charnber of spout and bridge bonles." The dichotomization of representation into these two categories supports the argument that Spondylus was an indicator of the status held by those members of Chimu society who controUed and ammged the mechmïsms of trade. The extensive arnount of Spo~lusremvered hmthe North Coast area indicates not only iacreased use of and demand for large supplies of the shell, but suggests that shell passed through coastal cultures to the cultures in the highlands. There is more evidence of concentrated Spondylus use on the north coast than any other area, so fm- The possibility of coast-highland exchange with Chan Chan as the coastal terminus is indicated by the excavation of Unit BB by John Topic (1977). The large number of llama burials in the unit, llama dung, possible sleeping platforms, a communal kitchen, and a macaw skeleton native to the eastem side of the Andes, indicates that Unit BB was a trading terminal where llarnas and the caravan dnvers were quartered (Topic 1977:275-281). However, while there is ample endence of Spondylus use at Chan Chan, there is no Spondylus evidence from this particular excavation that would tie in Spondylus as one of the aansported materiais. On the south coast, the role of Chincha as a centre of trade in the Late Horizon has already been discussed in Chapter Four, but 1 will review the archaeological distribution of Spondyim here. Max Uhle excavated several sites in the Valley and concluded that there were no Spondylus shells or artifacts in the pre-Inca bwials, only in those of the transitional and Inca phases (Kroeber and Strong 1965:30). Burials excavated in the Pampa de Canelo, in front of the ruins of La Centinela, included shells, necklaces, beads, pendants, and zoomorphic fonns, some of which were inlaid with green Stone (Kroeber and Strong 1965:3 0-3 1). Later excavations at Lo Demis in the Chincha region by Daniel Sandweiss (1 992: 102) tumeci up only a few Spondyius fragments in an area of large-scale fishing activity. ïhese finds do not necessarily substantiate the position of Chuicha in a maritime network, as many other coastal sites show evidence of the cultural use of Spondyiw h ho ut king tied to a trade network. Spodjdus artifacts cannot always be considered as indicators of commerd trade. The movement of goods can, at times, be nothing more than a case of simple transportation winthoutthe intent to trade. The discovery of red and white Spondyfusshell figtirïnes shapeà like llamas and humans at sites such as Cerro Copiapo in the Chilekgentine Cordiliera (kibarren Charlin 1978:447; Reinhard 1992:90) could be a result of the Inca expansion of Collasuyu in the Late Horizon and do not necessarily indicate that the region was previously part of a maritime trade network. Similariy, the Spondylus beads fiom the Calchaqui Valley. Argentina, appear in the production debris of elite Inca households, where the oficials in charge of relocating the local wealth to the Inca empire were stationed (Earle 1994). Military carnpaigns and expansion during the Inca conquest may also have redistributed finished Spondylus objects north away fiom Cuzco. Some of the Spondylus artifacts found in Late Horizon Ecuadonan sites might have been carried there after first being importeci to Cuzco. Shipment of Spondylus fiom the coast of Ecuador to the centre of the Inca empire was one of the functions of the mullu chasqui camayoc. As the empire expanded, objects of riaial value could have been camied to new locations. The llama figurine found in Chile and the llama figurine found in Cuenca at Puxnapungu (Ecuador) may have been carved in the same location and carried by an Inca official. In this sense, then, the recovery of a SponrrS,Ius figurine in the highlands of Ecuador for this time period need not indicate that Spondylus was arriving directly nom the Ecuadorian coast. The archaeological recovery of these artifacts should indicate the possibility that finished Spondylusartifacts might not aiways be found at the site of their initial destination, but instead could have been redistributed to secondary sites. The archaedogicaL evidence of Spondylus use presented above seems to support the various associations of mullu that were discussed in Chapter Three. However, on closer examination the evidence appears to corroborate some associations only within specific geographic areas and time periods. While the list of sites contained here is by no means complete, it does represent a cross- section of those sites within each area where the associations are related- The gender metaphor appears most strongly on the Phase D p!aque fiom Chavin de Huantar in Peru. and at sites with strong Chavin influence such as the Paracas culture and the Carhua textiles, but thete is littie evidence that the same association applies to the rest of the Andes. The intennent of Spondylus with female skeletons has been mentioned for Kuntur Wasi (Kato 1993:22î),La Galgada (Gneder 1988: 89?92,94). Pacatnamu (Bruce 1986:LOS), and Pachacarnac (Uhle 1991 [1903]:95), in Peru Apart from these connections there is little direct evidence elsewhere in the Andes of the female association. Males were also buried with Spondylus grave goods. The burials of female skeletons at Ingapirca by the Cariarï in the Late Intermediate Period near Cuenca with Spondylus may be nothing more than coincidence, but because of the Chavin information there is a tendency to relate al1 Spondylus with female characteristics. For the same reasons the place of Spondylus in dual opposition to Strombus is similady identified with evidence from the Early Horizon. Water associations corne more from historic records dian archaeological evidence. Spondylus sacrificed for water does not necessarily have to appear in the archaeological record as Spondylus anifacts contauied by water. Caches of Spondylus do not always imply that sacrifice for water was their reason for being, and yet the relationship of mullu to wter is seen as a common trait throughout the Andes. Much ofthis reiationshÏp is based on the myths of Huarochiri in the central highlands of P~Nand the taies of the "daughters of the sean by Cobo (1 990) and Acosta (1962). Simüarly, the suggestion that Spodylus was seen as food of the gods is not based on archaeological evidence but again stems fkom one source: the myths of Huarochiri. The use of Spondylus chaqrtira as a form of monetary wealth in the northem Andes and on the coast of Ecuador does not appear to have spread south to Pem where chaquira is found as dedicatory offerings in post holes and under floors. When Spondylus was used in a pst hole it did not necessarity engender that building with female characteristics, any more than it meant the building was dedicated to the acquirement of water A building with Spondylus interred under it may have held a rinial significance to the acquirement of water, but in other locations such as Chan Chan it appears to indicate rank or status (Day 1973:276). Perhaps the connection lies in the mler's responsibility to provide water for the general population. In these instances it can only be seen to show that the shell was an important ritual and symbolic item, highly regarded by the people who put it there. Those concepts rilated to Spondylus such as the SpondyludStrombus dyad do not appear to have been universal characteristics. At Cerro Narrio the shell was predominantly used in the manufimure of figurines that remaineci in the local area, although the site was on an east-west trading route from the coast to the Oriente. During the Middle Horizon at Huari sites Spondylus is used in dedicatory offerings and associated with turquoise. ïhe Chimu used whole? worked and crushed shells but the association is with Conus not Strombus. From the Late Horizon at Cabeza de Vaca (Tumbez) Spondylus is found in association with Smbus, Conus, and various other shells. While it appears that Spondyltls was not the ody Unportant shelt m Andean histoq, it can be seen to have been a coasistentiy important shell over both time and space. The mcrease in Spondylus use during the Middle Horizon can be seen archaeologically, but the reasans for its increase are iess cl=. One theory suggests that the increase rnay have been brought on by the adverse effects of a 32 year long drought that hit the Centrai Andes starting in AD 562. The shell may have been widely used in rituals to appeal to the deities responsible for the retum of water (Shimada 1991: LI). However, this reasoning conceming the increased use is resaicted to a single Spondylus characteristic, to bruig water, that is contradicted by its other archaeological contexts. The increase in use is more likely the combined result of its other uses. There are no easily identifiable characteristics that can be used to define Spondylus use in the Andes based on this evidence. The only common denominator for al1 these situations is that Spondylus was an important highly regarded and rnuch desired object. The archaeological evidence conceming Spondylus ïndicates the same conclusions as that discussed for nullu in Chapter niree. The multiple meanings of mullu are as consistent as the multiple archaeological contexts of Spondylus- A more definable trait for Spondylus is its appearance as a pre-Columbian trade material. Full-blown trade networks do not instantly appear but expand over time. Archaeological investigation has shown that the range and quantities of Spondylus use in the Andes increased over several thousand years. The methods of procurement and distribution indicate those mechanisms used in transporhg goods by land or sea, over long or short distances. nie movement of Spondylus and its appearance in the archaeological record is an important indicator of trade and exchange systems. It might also be argued that the lack of Spondylus in the archaeological record is just as important. For example, as important as the shell was duhg the Late Horizon to the Inca, it is strange that no evidence of its use appears at H~UCOPampa: an ancient Inca administrative and storage centre (Moms and Thompson 1985). Increased Spondylus use duaughout the chtonological sequence could be used to support the presence of a maritime trade network dong the coast of Ecuador and Peru, and possibly nordi to Mesoamerica Spandyllus trade may well have been the reason for the creation of a network, and one of the major commodities transported. However, the distribution of Spondylus at prehistoric sites such as Aspero, Chavin, Garagay, and La Gaigada indicates that Spondylus possibly appeared in Peru long before the evidence suggesting a maritime long- distance trade nehNork appears. If the addition of marine transport is viewed as nothing more than an improvement to the dready existing network of land-based trade, then the improvements are only in the area of quantity and speed of distribution. The archaeological distribution of Spondylus does little to reconcile the maritime-network debate. III

CHAPTER SIX

CONCLUSIONS

This thesis has examined the importance of mullu to Andean cultures. From chaquira beads to the vagina dentata, Spondylus and mullu represented both concrete and abstract elements of Andean cultures. At the basic level, mullu was simply a Quechua word used to refer to Spondylus shell in al1 its manifestations. In a more specific sense mullu was the chaquira beads made from the Spondylus shell. In its broadest meaning, mullu was an umbrella term that included other materials such as mother-of-pearl, turquoise and Strombus. Mullu evokes different relationships of materials at different levels and times depending upon the ritual need and its context. The concept of rnullu was a complex that included these other items, their ritual use, and trade value, representing multiple levels of cultural involvement at any given tirne. Mile serving as chuquira, mullu could also be used in rituai ceremony or as a personal adornment by the same culture poup. The possibility that mullu was simply a synonym for Spondylus must be discounted. In the Introduction 1 asked the question, "while Spondvlus is mullu, is mullu necessarily Spondylus?". to which the answer must be, "not aiways". It is apparent that Spondylus was used as mullu, in the sarne sense that chaquzra beads were mullu made of Spondylus, but it is equally apparent that mullu was much more than just Spondylus. The possibility that mullu was a term for a complex which included Spondylus and other items such as mother-of-pearl, Strombus, foodstuffs, and metaphorkal associations, desemes serious consideration. The work by Ann Mester (1990) on rn~ther-o~peaflas a fonn of mullu provides a foundation on which to question the materiai make-up of nzullu. Further investigation of the chronicles should increase the inventory of possible mullu materials which appear to Vary in colour and substance. If we accept that mullu repmnts different mater& or combuiations of materials then we must aiso accept that its meanhg changes depending upon the ritual need and its contes We know fiom the chronicles that mullu was used in sacrifice, but was Spondylus referred to as somethïng else when used in dedicatory offerings under floon and in post holes? The mullu used in ceremonial sacrifice appears in combination with other materials (Guaman Poma 1980; Molina 1989 [1575]: 133), similar to the combinations called mesas used by modem healers. Does the combination of mulh with other sacred objects increase or alter its sacred power, or can the power of mullu be imbued on another object? If there are other materials called mullu, such as foodsniffs then we must also deal with materials that are mullu-like, or related to the concept of m ullu. Finally, if the concept of mullu did include these other items, their ritual use, trade value, and sumpniary use, then it was being used on multiple ievels and as part of what could be considered as a ritual and trade complex. Through its position in the complex. and very likely at the centre of it, Spondylus becarne associated with the methods of procurement and distribution that developed into long-distance trade networks. While the primary uses of nullu are of interest to the study of socio-cultural belief and titual, the secondary characteristics of Spondylus as a valued nade item provide information of a socio-economic nature, However, the long-distance exchange of Spon&lus should not be seen as a monocausal expianation for the formation of trade networks. Rather it was one of many commodities travelling tbn,ugh the networks. While Spondylus could only be procureci through long-disiance trade, there were other materials nquired by Andean groups which could have necessitated the fornation of a network. References to the importance of pari oyster sheII, arsenical copper production, and the trade of technicd knowledge ali ùidicate that other materials were moving through the networks, not to mention the short-distance trade in subsistence and utilitarian goods. With its role in Andean cosmology and its association with trade networks, Spondylus can be seen as an object of extremeiy high value. With its many diEerent representations within the cosmology of the Andes nullu can be seen as a concept that forms the basis for a rituai and trade complex of characterktics, In the past twenty years, Spondylus has become an important aspect of Andean archaeology. The results of my thesis indicate, that it is now time for a more complete investigation into the use and concepts behind its alter-ego, mullu- Abbott, R Tucker 1974 Ancericon Seashelk The Mrîne MolIusca of the Atlantic and Pacifc Cousis of Norlh Arne&, Second Edition. Van Nosirand Reinhold Company, New York.

Acosta, Joseph de 1%2 [ISW] Historia Nafural Y Moral De Las India, 2nd edition. ed. by Edrnundo OIGoman.Fondo De Cultura Eoan~mica,Mexico.

Adams, Robert McC. 1974 Anthropological Perspectives on Ancient Traie. Currenf Anthropology 15239-258.

Adorno, Rolena 1982 The Language of History in Guaman Pomds Nuevu CorOnica y Buen Gobiemo. ln FmOral to Wntten mression: Native Andean Chronicles of the Eariy Colonial P eriod, ed. by Rolena Adorno, pp 109-i 73, Maxwell School ofcitizenship and Public ARairs, Syracuse University, Syracuse.

Albornoz, Crïstobal de 1989 [iSSsJ I~ccionPm Descubrir ThLas Guacas Del Piru Y Sus Camayos Y Hm*en&s. Ed. by Henrique Urbano and Pierre Duviols. Madrid, Spain.

Allan. Richard n-d. Site OMJPLP MO: A Speciaiist Shell Workshop (ca 1989). MS.

Allen. Catherine J. 1988 The Hold Life Hm Smithsonian Institution, Washington.

Alva Walter 1988 Discovenng die New World's Richest Unlooted T'omb. Notional Geogruphic l74(4):5lM49.

Anawalt, Pamcia Rieff 1992 Ancient Cultural Contacts Between Ecuador, West Mexico, And The American Southwest: Clothing Similarities. Latin American Antiquiîy 3: 1 14-129. Anders, Manha B. 1% WM Experiments in Staîecdk A View firom Azangaro. In, Andean Archaeologyr Papers m Memory of CIford Evris, ed. by Ramiro Matos M., Solveig A. Turpin, Herbert H. Elîag, Jr., pp 20 1-224. University of California, Los Angeles.

Amaga, Pablo joseph de 1%8 [1621] The Extirpation Ofldolrioy In Penr Translated and edited by L. Clark Keating. University of Kentucky Press, Lexington.

Athens, J. Stephen 1992 Ethnicity and Adaptation: The Late Period-Cara Occupation in Northem Highiand Ecuador. In, Resources, Power, and Interregional Interactionyed. by Edward M. Schortman and Pamcia A. Urbm pp 193-219. Plenum Press, New York.

Avila, Francisco de 1975 [lS98] Dioses y honrbres de Hzuzrochiri Second Edition translation by José Maria Arguecias, ed. by Pierre Duviols. Sig10 Veintiuno Editores, SA, Mexico.

Bach, Robert L. 1980 On the Hoiism of a World-Systems Perspective. In, Processes of the WorldSysem. ed. by Terence K. Hopkins and Immanuel Wdlemein, pp 289-3 10. Sage, Beverly Hills.

Bandelier, Adolph F. i%9 [1910] The Islands of Titicaca und Koati. Kraus Rep- New York.

Bird, Junius B. 1943 Excavatiom In Northem Chile- The American Museum Of Natural History, New York.

Bird, Juntus B. and John Hyslop 1985 The Precermnic Excuvations ut the Huaca Prieta Chicmu Valley. Pen. American Museum of Natural History, New York.

Bonavia, Duccio 1982 Los Galivones: mm. desierto y obis en la historia del hombre. Editorial Ausonia-Tallers Graficos S. A., Lima. Borah, Woodrow 1954 Eàr& Colonid Ti& And Nmgation Between MicoAnd Pm Ibero-Amencana: 38, Univemty of CaIifoCLUa, Berkeley.

Bourget, Steve 1995 The Chilben of the Mountmn d the Llàntus of the Ancestors: Funermy and Feralty Ririurs a? Fbe Midde Horikon Sites of the Moche Valley, Pem Paper presented to the 35th Annuai Meeting of the InstiMe of Andean Studies, Berkeley California.

Bray, Tamara L. 1992 Archaeologicai Smey In Northern Highland Ecuador: Inca Imperialisrn and the Pais CaranquÏ. World Archaeology 24:218-233.

Bmce, susan Lee 1986 The Audiencia Rom of the Huaca 1 Complex. In, The Pacatnamu Paps Volme 1, ed. by Chnstopher B. Doman and Guillermo A. Cock pp 95-108 University ofCalifornia, Los Angeles.

Bruhns, Karen Olsen 1989 Intercambio entre la costa y la sierra en el Fonnativo Tardio: Nuevas evidencias del Azuay. In, Relaciones interczilturales en el hrea enrotorial del Phciifico durante la épma precolombina, ed. by L F. Bouchard and M-Guinea pp 57-74. BAR International Series 503, Mord-

1994 Ancient South America- Cam bridge University Press, Cambridge.

BrumfieI, Elizabeth M. and Timothy K. Earle 1987 Specialization, Exchange, and Complex Societies: an introduction. In, Speciukation, Exchange, and Complex Societies, ed. by Elizabeth M. Brumfiel and Timothy K Earle, pp 1-9. Cambridge University, Cam bridge.

Burger, Richard L. 1984 The PrehistorÏc Occtlpation of Chovin de Huhtur, Peru. University of California Press, Berkeley.

1992 Chmin And The Origins Of Andean Civilùation. Tharnes and Hudson, London. Burger, Richard and Lucy Salazar-Burger 1993 The Place ofDual Organization in Eariy Andean Ceremonialism: A Comparative Review. In, El MdoCeremonial Andino ed. by Luis Miliones and Yoshio On* pp 974 16. Semi Ethnologicai Studies 37, Osaka.

Burger, Richard L., Frank Asam, Helen V. Michel, Fred H. Stross and Emesto Salam 199% An Initial Consideration of Obsidian Procurement and Exchange in Rehispanic Ecuadorf Lam American Antiq* 5(3)228-255.

BushneU, G.H. S. 1951 me Archaeology Of The Santa Mena Peninda In Sbuth-West Ecuador. Cambridge University Press, Cambridge.

Buys, Jozef E. and Victoria Dominguez 1989 La Arqueologia De Cumbaya, Provincia De Pichincha, Ecuador. In, Relaciones interculturales en el &eu ecuatonol del Phj?co durante la época precolombina, ed. by J. F. Bouchard and M. Guinea pp 75-95. BAR International Series 503, Mord.

Cabello Valboa Miguel 1% 1 [1 S8qMiscelhea Antbrtrrtrca:Una Hiauria Del Peni Antiguo. Universidad Nacional Mayor De San Marcos, Faculdad De Lem, Instituto De Etnologia, Lima

Carcedo Muro, Paloma and hiShimada 1985 Behind the Golden Mask: Sicin Gold Amfacts fkom Bath Grande, Peru. In, The Art of Precolwnbian Gold: The Jan Mitchell Collection, ed. by Julie Jones. Little, Brown and Company, Boston.

Camuiza, Martin de 1965 [l%9] Relacion De Las Provincias De Las Esmeraldas Que Fue A Pacificar El Capstan Andrés Contero. In, Reluciones Geogr@cas de Indias. ed. by Marcos Jiménez de la E@a, Vol. 3 pp 87-90. Ediciones Atlas, Madrid (Biblioteca de Autores Espalloles Vol. 185).

Carrasco, Pedro 1983 Some Theoretical Considerations About the Role of the Market in Ancient Mexico. In, Economic Anthropology: Toprcs and Kheories, ed. by Sutti Ortiz, pp 67-82. University Press of Arnenca, Lanham. C-n Cachot, Rebecs 1949 Paracar Cultural Element~.Corporaci6n Nacional de Turismo, Lima

1955 El CuIto Al Agua En El Antiguo Peru: La Pcrccha elemento cultural Pan-andino. Sepamta de la Revista del Museo Nacional de Antropogia y Arqueologia, Lima

Claasseti, Cheryl and Samuella Sigmann 1993 Soming Busycon Artifacts of the Eastern United States. Amerkm Antiquiîy 58 (2)pp. 333-347.

Clark,Grahame 1986 Symbols of Excellence: Precious materials as expressions ofstaîus. Cambridge University Press, Cambridge.

Cobo, Bernabé 1979 History of the Inca Empire. Translated and edited by Roland Hamilton. University of Texas, Austin.

1990 Inca Religion and Customs. Translated and edited by Roland Hamilton. University of Texas Ress, Austin.

Collier, Donald 1955 Cultural Chronology and Change as Reflected in the Ceramics of the Virii Valley, Pem. Fieldianu: Anthropology Volume 43. Chicago Natural History Museum, Chicago-

Collier, Donald and John V. Murra 1943 Survey And Excavations In Southem Ecundor. Field Museum of Natural History ,Chicago.

Conrad, Geofftey W. 19û2 The Burial Platforms of Chan Chan: Some Social and Political Implications. In Chan Chan. Andean Desert City, ed. by Michael E. Moseley and Kent C. Day, pp 87-1 17, University of New Mexico, Albuquerque.

Cook, Anka G. 1992 The Stone Ancestoa: Idioms Of Imperia1 Attire And Rank Among Huari Figurines. Latin Amencan Antiquity3(4):341-364 Cordy-Collins, Aiana 1978 The Dual Divinity Concept In Ch- art El Dorodo IIi(2): 1-3 1.

1990 Fonga Sigde, Shell Purveyor to the Chimu Kings. In, 7ne Northem Dynasties: Kingship and Stutemsft m Chimor, Michael E. Moseley and Nana Cordy-Collins editorst pp 393-417. Dumbarton Oaks Research Library and Collection, Washington, D.C..

Cowgill, George L. 1979 Teotzhuacan, Interd Militaristic Cornpetition, and the FaIl of the Classic Maya In, Mqya Archaeology and Ethnohistory, ed. by Norman Hammond and G.R Willey, pp 5 1-62. University of Texas, Austin-

CraigyAlan K. and Norbert P. Psuty 1968 The Pmacos Pupers: S.esIn Marine Desert Ecology Vol. 1, No. 2, Reconnaissance Report. Florida Atlantic University, Boca Raton.

Davidson, Judith R 1980 The Spon&ius Shell In Chimu Iconogruphy. Master's Thesis. California State University, Nordiridge.

1981 El Spondylus En La Cosmologia Chirnu. Revista EI Museo NactonaI. Tomo XLV: 75-87-

Day, Kent C. 1973 Architecture of Ciuddela Rivero, Chan Chan. Peru. Unpublished Ph-D. Dissertation. Harvard University, Cambridge.

Donnan, Christopher B. and Carol J. Mackey 1978 Ancient Burial Patterns of the Moche Valley. Pem. University of Texas, Austin.

Doman, Christopher B. and Leonard J. Foote 1978 Appendix 2: Child and Llama Burials nom Huanchaco- In, Anczent Burial Patterns of the Moche VaIky. Pen, pp 399-402. University of Texas, Austin. Dorsey, George 1901 ArchaeologicaZ Imesfrgatioms on the Isid of Lu PIaa EMUlCjlOr- Field Museum of Naiural History, Chicago.

EarIe, Timothy K. 19W Wealth Finance In The Inka Empire: Endence From The Calchaqui VaUey, Argentins Amertean Antiquity59(3):4431460-

Edwards, Clinton R 1%9 Possibilities of Re-Columbian Maritime Contacts Among New World Civilizations. In, Precolumbian Con- Within NucIear Americu, ed. by J. Charies KeUey and Carroll L. Riley, pp 3-10. Southem Illinois University, Illinois.

lm New World Perspectives on Re-Europzan Voyaging in the Pacific. In, EdyChinese Art And Its Possible IMuence in the Pucifc Basin. Volume ïïwee, Oceania and the Amencas. Ed. by Noel Barnard pp 843-887. Intercultural Arts Press, New York.

Elera Carlos G. 1993 El Cornplejo Cultural Cupisnique: Antecedentes y Desarrollo de su Ideologia Religiosa. Ln, El Mundo Ceremonid Andino ed. by Luis Millones and Yoshio Onuki, pp 229-257. Senri Ethnoiogicai Studies 37, osaka

Engel, Freàeric 1%3 A Preceramic Senlement on the Central Coast of Peru: Asia, Unit 1. In, Transactions of the Ametican Philosophical Society. Nèw Series - Volme 53. Pm3. The Amencan Philosophical Society, Philadelphia.

Enloe, Cynthia H. 1980 State-Building and Ethnic Structures: Dependence on International Capitdisî Penetmtion. In, Processes of the World-System. ed. by Terence K. Hopkins and Imrnanuel Wallemein, pp 266-288. Sage, Beverly Hills.

Espinoza Galana, Max 1973 Topbnimos Quechuas Del Pem. Cose- Lima.

Feldman, Lawrence H. 1972 Shells From Afàr. 'Panamic' Molluscs In Mayan Sites. In, Mesoamerica Archaeology: New Approaches. ed. by Norman Hammond, pp 129- 133. University of Texas, Austin. Feldman, Robert A. 19&2 -c Corpotate Archifecture: Evidence For ihe Development of Non-EgaiiiEaran Social Systems in Peni. In, EwZy cerenomal Architecture h the Andes, d by ChrïstopherB. Donnan pp 71-92. Dumbarton ûaks, Washington D-C..

1992 Preceramic Architectural and Subsistence Traditions. in, Andean Pust Vohme 3, ed. by Daniel H- Sandweiss, pp 67-86, Comell University, Ithaca.

Fonseca Z., Oscar and James B. Richardson ïü 1978 South Amerïcan And Mayan Culturai Contacts At The Las Huacas Site, Costa Rica Adsof Carnegie Mweum47(13):29%317.

Fomee, Niculoso de 1%5 [1586] Descrïpcion de la tiem del corregimiento de Abancay, de que es corregidor Niculoso de Fomee. in, Relaciones geogv@cus de Indics- ed. by Marcos Jiménez de la Espada, Vol. 2 pp 1630. Ediciones Atlas, Madrid (Biblioteca de Autores Espafioles Vol- 184).

Galvh Garcia, J. R and M. A. Barrius0 Pérez 1986 Estudio de las chaquiras de Atacames (Ecuador). MisceIhea Anîropol6gica Ecuatoriunu6: 6 1-80.

GBord, Douglas and Pauline Hoggarth 1976 Cumival and Coca Le@ Sme Traditions of the Peruvian Quechua Ayllu. St. Martin's Press, New York.

Gonçalez Holguin, Diego 1952 [1608] Vocabulario De La Lengua Generd De Todo El Peru LLamada Lengua QQuichua O Del Inca. ed. by Rad Porras Barrenecka-Edition Del Instituto De Historia, Lima

Grieder, Terence 1978 nie Art and Archneology of Pmhash. University of Texas Press, Austin.

1988 Burial Patterns and Offerings. In, La Galgacla. Peric A Precermnic Culhrre In Transition. pp 87-95. University of Texas Press, Austin. Grïeder, Terence and Alberto Bueno Mendoza, C. Earle Smith and Robert M. MaIina 19ê8 La GulgaclO. Pex A Precerumic Culture In Transition. University of Texas Press, Austin.

Guaman Poma De Ayala, Feiipe 1980 El Primer Numa Coronica Y Buen Gobierno. Ed. by John V. Murra and Rolena Adorno. Siglo Veintimo, Mexico.

Guinea, Mercedes 1989 Valoracion de las evidencias de intercambio en la Desembocaàura del rio Esmeraldas: El problema cronologico. In, Relaciones intercuitamales en el &ea ectlotorial del POcifico dwmte la époco precolonzbina, ed. by J. F. Bouchard and M.Guinea pp 12%146. BAR International Series 503, Oxford.

Haas, Jonathan 1985 Excavations on Huaca Grande: an Initial View of the Elite of Pampa Grande, Peru. Journal of Field Arckeology lZ(4): 39 1-410.

Halstead, Paul 1993 Spondy lus shell ornarnents fiom late Neolithic Dimini, Greece: specialized manufacture or unequal accumulation. Antiquity67: 603- 609.

Harbottle, Garman and Phi1 C. Weigand 1992 Turquoise in Pre-Columbian Amenai. Scient@ Antericm 266:7885.

Harrison Regina 1982 Modes of Discome: The Rekucibn de antigüedades deste reyno del Pini By Joan De Santacniz Pachacuti Yamqui Salcamaygua. In From Oral 20 Written Expression: Native Andean Chronicles of the Em Colonial Period, ed. by Rolena Adorno, pp 65- 100, Maxwell School of Citizenship and Public -airs, Syracuse University, Syracuse.

1989 Signs, Songs. and Mernos, in the Andes. University of Texas Press, Austin.

Healy, Paul F. 1990 Excavations at Pacbitun, Belize: Preliminary Report on the 1986 and 1987 Investigations. Journal of Field Archaeology 17(1):247-262. Healy, Paul F., Heather 1. McKillop and Bernie Walsh 1584 Analysis ofObsidian fÎom Moho Cay, Belize: New Endence on Classic Maya Trade Routes-Science225414-416.

Hem, César M. and Martinez l& Ad- Pavon 1989 Chimeneas Ceramicas: Un Rasgo Culturai De Signincacion Controvertidac In, ReIaciones intermllriaoles en el &ea ecuatoriaf del Phc@co durante la época precolombma, ed. by J. F. Bouchard and M. Gu- pp 147-162. BAR Intemational Senes 503, Oxford.

Herrera, F. L. and Eugenio Yacovleff 1935 El Mundo vegetai de los antiguos peruanos (conciusion). Revisra Del Mweo Nacional4:3 1- 102,Lmia

Heyerdahl, Thor 1957 Guam Navigation: indigenous Sailing OEThe Andean Coast. Southwestern Journal of Anthropology 1 1: 25 1-264.

Heyerdahl, Thor and Anie Skjelsvold 1990 Archaeologica2 Evidence of Pre-Spanish Visits to the Galapagos Islands. Norwegian University Press,Oslo.

Heyerdahl, Thor and Daniel H. Sandweiss and Aifiedo Narvaez 1995 Pyramt& of ï'ricume: The Quest Fur Peru's Forgotten City. Chapter 1, Thames and Hudson, London.

Hocquenghem, Anne-Marie 1991 Frontera Entre "Areas Culturales" Nor Y Centroandinas En Los Valles Y La Costa Del Extrerno Norte Periiano- Bulletin de I'lnstirut Francais drEtudes Andines20: 309-348.

1993 Rutas De Entrada Del Mullu En Ei Extrerno Norte Del Peni- Bulletin de 1Ynstitut Francois d'Emdes Andines23(3):70 1-719.

Hocquenghem, Anne Marie and Manuel Pefia Ruiz 1994 La Talla Del Material Malacologico En Tumbes. Bulletin de l'institut Francais d Etudes Andines23(2):209-229.

Hodder, lan 1978 Some Effects of Distance on Patterns of Human Interaction- In, The Spatial Org~nisationof Culme, ed. by ian Hodder, pp 155- 178. University of Pittsburgh, Pittsburgh. 1983 The Present Past. Pica, New York.

Holm,Olaf 1%6/67 Money Axes firom Ecuador. Folk: Dan& Emogrisk Tidsskn'i VOL-8-9:135-143.

1% 1 Culnna MiIagro-Q~evedo~Banco Centrai Del Ecuador, Guayaquil.

Hong, Suk Ki, Donald W. Rennie and Yang Saeng Park 19û2 Cold Acclima~utionand DeaccIimutiiafio~of Korean Women Divers. Unpublished rnanuscrip& Department of Physiology, State University ofNew York at BuffbLo, New York.

Hosler, Dorothy 1988 Ancient West Mexican Metallurgy: South and Centrai Amencan Origins and West Mexican Transformations. American Anthropologist 90: 832-855.

Hosler, Dorothy and Heather Lechtman, and Olaf Holm 1990 The Technologies of Exchange. In, Axe Monies and nieir Relatives, pp 70-102, Dumbarton Oaks, Washington.

Hyslop, John 1984 The Inka Rood System. Academic Press, Orlando.

Instïtuto Geografico Militar n/d Cndice Toponimico de la Repblic del ~cuador,WI(LL-M-N-13-0). iribarren Chariin, Jorge 1978 Manifdons of Inca Culture in Two Provinces of Chiie. In, Advances in Andean Archaeology, 4-by David L. Browrnan, pp 443- 452. Aldine, Chicago.

Isbell, Billie Jean 1978 To Defnd Ourselves: Ecdogy and Ritual in an Andean Village. University of Texas, Austin.

Jones, Julie 1979 Mochica Works of Art in Metal: A Review. in, P re-Columbimi Metallurgy Of South America: A Conference ut Dumbarîon Oak, October 18th And 19th, 1975. Ed. by Elizabeth P. Benson, pp 53-104. hunbarton Oaks Research Lïbfary and Collections, Washingîon. Karsten, Rafael L926 The Civiiization of the South American Indans: With Spectaf Reference to Magic and Religionn Alfred A. Knopf, New York.

Kato, Yasutake 1993 Resultados de las excavaciones en KUntur Wasï, Cajamarca In, Ei rnundo ceremofifalandino ed. by Luis Millones and Yoshio Onuki, pp 203-228.SenrÏ Ethnologicai Studies 37, Osaka

Keatinge, Richard W. 1982 The Chimii Empire in a Regionai Perspective: Cultural Antecedents and Continuities. In, Chan Chmc Andean Desert Ci@,ed. by Michael E. Moseley and Kent C. Day, pp 197-224, University of New Mexico, Albuquerque.

Keen, A. Myra 19'71 Sea Shells of Tropical West Amerka: .Marne Mollurks fion Baja Caizjiornia to Pem 2nd ed., Stanford University Press, Stanford.

Kohl, Philip L. 1981 The Use and Abuse of World Systeïns Theory: The Case of the Pristhe West Asian State. In, Advances in Archaeoiogicai Method and Theory Volwne 1 1, ed. by Michael B. Schiffer, pp 1-35. Academic, San Diego.

Kolb, Charles C. 1987 Marine Shell Trade and Classic Teotihuacan. Mexico. BAR tnternational Series 364. Oxford, England

Kroeber, A. L. 1%5 The Uhle Pottery Collections From Moche. In, (Iniversity of Cakjiornia Publications in Amencan Archaeoiogy And Ethology, Volume XU 192401927, ed. by A. L. Kroeber and Robert H. Lowie pp 19 1-234. Kraus Reprint Corporation, New York.

Kroeber, A. L., and William Duncan Strong 1x5 The Uhle Collections fiom Chincha In' University of Culifornia Publications in Americtm Archaeology And Ethnology, Volume XXI 1 924-1 92 7, ed. by A. L. Kroeber and Robert H. Lowie pp 1-54. Kraus Rep~tCorporaîion, New York.

Lamprell, Kevin 1987 Spondylus: Spiny ester Shells of the World E. I. Brill, Leiden. Lam, IesUs 1978 Dicciottario Qhëshwa-Cc~steiIÙnoCasteIlano-Qhëshwa segundu edicibn. Los Amigos del Libro, La Paz

Larco Hoyle, Raniel 1%5 Checan: Essai Sur Les Représentations Érotiques Du Pérou Précolombien. Les Éditions Nagel Genéve.

Lastra, Yolanda 1%8 Cochabamba Quechua S'.Mouton, Paris.

Lathrap, Donald W. 1973 The antiquity and importance of long-distance made relationships in the moist tropics of Pre-Colurnbian South America World Archaeology 5:170-1 86.

1985 Jaws: the Control of Power in the Early Nuclear American Ceremonial Center. In, EarIy Ceternonial Architecture in the Andes. d.by Christopher B. Donnan, pp 24 1-267,Dumbarton Oaks, Washington.

Lathrap, Donald W., Donald Collier and Helen Chandra 1975 Ancient Ecuador: Culture. CIay and Creativity 3000-300 B. C. Field Museum of Natural History, Chicago.

Lehman, Walter 1975 [1924] The Art Of Old Peru. Eiacker Art Books, New York.

Lopez De Velasco, Juan 1971 Geogr@a Y Description Universal De Las Indias* ed. by Marcos Jiménez De La Espada. Atlas Ediciones, Madrid.

Lumbreras, Luis Guillermo 198'7 Childe And The Urban Revolution: The Centrai Andean Experïence. In, Sudies In The Nedithic And Urban Revohtions, edited by Linda Manzanilla pp 327-344. BAR International Series 349, Oxford.

Lynch, Thomas F. 1989 Regional Interaction, Transhumance, And Verticality: Archaeological Use UfZonal Complementarity In Perii And Northem Chile. In, Michigart Dise- in Anthropology 8: 1 - 1 1, University of Michigan, Am Arbor. Lyon, Pamcia J, 1978 Female Supernaairas In Ancient P~Nkwpa Pacha 16: 9S-lM.

Mackey, Carol J. 19û7 Chimu Admini-on m the Provinces. In, Originr and Developnrent of the Andean Slate, ed. by Jonathon Haas, Thomas Pomrski, and Sheiia PozorskÏ, pp 121429. Cambridge University Ress, New York.

Mannheim, Bruce 1991 The Language of the Inko sinee the European Im>aron. University of Texas Press, Aue.

Marcos, Jorge G. 19'77478 Cnii-singto Acapulco and Back with the Thorny Oyster Set: A Mode1 for a Luid Exchange System. Journal of the Steward Anthropologicai Society 9:99-1 32. -

1986 Interaunbio a larga distancia en América: el caso del Spondyïus. in, Arqueologia De La Costa Ecuc~toriana:mevos enfoques. Jorge Marcos editor. pp 197-206, Centm de Estudios Arqueologicos y Antropologicos, Guayaquil.

Marcos, Jorge G-,and Presley Norton 198 1 interpretacion sobre arquelogia de la Ida de La Plata Miscelhea Anhopolt5gica Ecuutoriana. Boletin de los Museos del Banco Central Del Ecuador 1 :136-154.

McEwan, Colin and Maria Isabel Silva 1. 1989 'Que Fueron A Hacer Los incas En La Costa Central Del Ecuador? In, Relaciones interculturaïes en el drea ecuatoriai del Phczjko durante la época precolombïna, ed. by S. F. Bouchard and M. Guinea pp 163- 185. BAR International Series 503, Odord.

McEwan, Gordon F. and D. Bruce Dickson 1978 Valdivia, Iomon Fishermen, And The Nature Of The North Pacific: Some Nautical Pmblems With Meggers, Evans, And Estrada's (1965) Transoceanic Contact Thesis. Anzerican Antiquity43(3):362-371.

Meggers, Betty L, ClEord Evans, and Emilio Estrada 1%5 Eariy Fottnative Perioà of Co~stalEauador: The Valdivia and Machalilla Phases. Smithsonian Institution, Washington. Meighan. Clement W. 1%9 Cultural Simiiarities Between Western Mexico and Andean Regions in, Precoldian Contact Within rVuciear Amertca, eded.by J. Charles Kelley and CarroU L. Riiey, pp L 1-25. Southem Illinois University, iIlmois-

Memel, Dorothy 1%s New Data on the Huari Empire in Middle Horizon Epoch 2A. RmPa Pacho6:47-114-

19ï7 The Archaeology of Ancient Pen und the Work of Mm me. University of California, Berkeley.

Mester, Ann 1985 Un Taller Manteno de Concha Madre Perla del Sitio Los Frailes, Manabi. Misceihea Antropologica Enratorima 5: 10 1 - 1 12.

1989 Marine Shell Symbolism in Andean Culture. In, Pmeedings of the 1986 Shell Bead CoMerence: Selected Papers, ed. by Charles F. Hayes III, pp 157-1 67. Rochester Museum and Science Center, Rochester.

1990 The Pewl Divers of Los Frailes: Archueofogical and Ethnohistoricol Expioratiom of Sumptumy Good Ti-ade and Cosnology in the North Andes. Ph-D. Dissertation, University of Illinois at Urbana-Champaign. University Microfilms International, Ann Arbor.

Moholy-Nagy, Hattula 1963 Shells And Other Marine Material From Tikal, Estudios De Cultura Maya 3: 65-83.

1989 Formed Shell Beads fkom Tikal, Guatemla In, Proceedings of the 1986 SM1 Bead Confrence: Selected Papers, ed. by Charles F. Hayes III,pp 139- 156. Rochester Museum and Science Center, Rochester.

Molina Cristobal de i9û9 [157qRelaci& de lus fabulas y ritos de los Inca. Ed. by Henrique Urbano and Piem Duviols. Madrid, Spain.

Moms, Craig and Donald E. Thompson 1985 Huhuco Pampa: An Inca City and its Hinterland. Thames and Hudson, London. Murra. John V. 1%3 The Historic Tribes of Ecuadot. In, Hdookof South American Indim. Volume 2 The Andean Civilizations, ed. by Julian H. Steward pp 785-82 1. Cooper Square, New York.

1975 [1971] El trafic0 de muüu en iacosta del Pacifico. In, Fornaciones Econbmics Y P olri'icas Del b.lundo Adho, pp 255-268. tnsîïîuto De Estudios Perhnos, Lima

Munia,Maninde 19û7 [1590]HistoriaGeneral Del Peni. ed. by Manuel Ballesteros Gaibrois. Historia 16, Madrid.

Navarro Santander, Hugo 1986 Proyecto Tacuynmo: Informe Final de Excavaciones. Centro de Documentacion y Estudios Arqueologicos?Tm. ilIo.

Netherly, Patricia J. lm Local Level Lorcis on the North Coast of Peni. Ph-D. Dissertation Comell University, University Microfilms International, Am Arbor.

Norton, Presley 1986 El seiiorio de Salangane y la liga de mercaderes: El cartel Spondylus- Baisa. in, Miscelh Antropok@ca Ecuatoriana .6 (6):1 3 1- 143.

Norton, Presley and Richard Lunnis and Nigel Nailing 1983 Excavaciones en Salango, provincia de Manabi, Ecuador. In, lbiisceianea Anîropologica Eciratoriana ,3(3): 9-72.

Olsson, Axe1 A. l%l Mollusks Of The Tropical Eastern Pacifie. Paleontological Research institution, Ithaca.

Parker, Gary John 1%9 Ayacucho Quechua Granimm And Dictiort~ry.Mouton, Paris.

Paul, Anne 1990 Paracar Ritual Attire: S'bok of Authority in Ancient Pen. University of Oklahoma Press, Norman. Pauisen. Allison C. 1%8 A Middle Horizon Tornb, Pinilia, Ica Valley Peru. Nmpu Pacha6: 1-6.

1974 The Thomy Oyster and the Voice of God: Spondylus and Stmmbus in Andean Prehistory. Amen'can Antiquity 29(4):597607.

1977 Patterns of Trade Between South Coastal Ecuador and Western Mesoamerica, 1500 B.C. - AD. 600. In, The Sen in the Pre-Columbian World, ed. by Elizabeth Benson pp 141- 160. Durnbmon Oaks, Washington D.C..

Pérez Bonay, Alfonso 1975 La Pesca En El Peni PrehispOnicu. Editorïals Unites SBA, Lima

Petersen G., Georg i97û Mineria Y Metalurgia En Ei Antiguo Pem. In, Arqueoiogicas, pp 24-26. instituto De Investi-gacionesAntropologicas, Lima-

Pires-Ferreira ,Sane W. 1982 Shell and Iron-Ore Mirror Exchange in Formative Mesoamerica, with Coments on Other Commodities. In, The Eudy Mesoamericon Village, ed. by Kent V. Flannery, pp 3 11-328. Academic Press, New York

Pillsbury, Joanne 1993 Sculped Frkzes of the Empire of Chimor. Unpublished Ph.D. Dissertation, Columbia University.

Ponce de Leibn, Sancho Paz 1%5 [1582] Relacion y description de los pueblos del partido de Otavalo. In, Reluciones geogrcificas de Indias, ed. by. Marcos Jiménez de la Espada, Vol. 2 pp 234-241. Ediciones Atlas, Madrid.

Ponas G.,Pedro 1. 1978 Arqueologia De La Cueva De Los T'os- Pontifica Universidad Catolica Del Ecuador, Quito.

1987 Nuestra Ayec Mmdde Arqueologia Ecw~toriiano.Centro de Investigaciones Arqueologicas, Quito. Pro* Donald A. 1985 An Analjwts of the Emiy Culmai Sequence in the Nep- Valley, Peru. Research Report Ntunber 25- Depamnent of Anthropology University of Massachusetts, Amherst

Puga Arroyo, Nicolis 1971 Toponimias Quechuas de Cajmarca y Sis Derredores. Editoriai Rumitiyana, TrujiUd=ajamarca

Purchon, R D. 1968 The Biologv Of The Mollwca- Pergamon Press, Odord.

Quiiter, Jeffiey 1985 Architecture and Chronology at El Paraiso, Pem. Journal of FieM Archaeology 12(3):279-298.

1989 Life and Death at Palontu: Society and Mortuary P ractices in a Precermic Peruvim Villoge. University of Iowa Press, Iowa City

Ramirez-Horton 1982 Retaïners of the Lords or Merchants: A Case of Mistaken Identity? Senri Efhnological Studies 10: 123-1 36.

Ravines, Rogger and Karen Stothert 1976 Un entierro comiin del Horizonte Tardio en la costa central del Peni. Revista Del Mirseo Nacional Tomo XLII: 153-206-

Real Academia Espailola 1976 117261 Diccionario De La Lengua Castellana Diccionurio De Autoridades: Edici6n Facsimil AC Editorial Gredos, Madrid.

ReicheCDolmatoff, Gerardo 1971 Aniaronian Cosmos: The Sexucti and Religow Symbolism of the Tukmo Indians of Colombia. University of Chicago Press, Chicago.

Reinhard, Iohan 199L Sacrd Peak of the Andes. National Geogrrphic, vol. 18 1 :84-1 12.

Richards, Horace G. and H. .J. Boekelman 1937 Shells From Maya Excavations In British Honduras. Amen'can Anîiquity 3: 166-169. Richardson, James B. 111, Mark A. McConaughy, Allison Heaps de PeiTa, and Elena B, DécimaZamecnik 1990 ïhe hiordieni Frontier ofthe Kingdom of Chimor: The Piura, Chira, and Tiimbez Vaiieys. In, The NonhDynusties: Kingship and Statecr&t in Chimor. Michael E. Moseley and Alana Cordy-Collins eds, pp 4 19-445- Dumbarton Oaks, Washington.

Rodriguez de Sandweiss, Maria del Carnien 1993 Malacologicd Andysis. In, Cahwchi in the Arzcieie World, by Helaine Siiveman, pp 294-299. University of Iowa, Iowa city.

Rollins. Harold B, James B. Richardson III, and D H. Sandweiss 1986 The Birth of El Nifio: GeoarchaeoIogical Evidence and Implications. Geoarchaeology 1 :3- 15.

Rostworowski de Diez Canseco, Maria Mercaderes del valle de Chincha en la época prehisphica: un document0 y unos cornentarios. Revista Esmolu de Antropologià Americana, 5: 135- 184, Madrid.

Pescadores, artesanos y mercaderes costefios en el Peni prehisphico. Revisfa EL Museo NaczonuI,Tomo XLI: 3 1 1-350.

Coastal Fishermen, Merchants, and Artisans in Re-Hispanic P~NIn, The Sea In The Pre-CoItunbiim World Elizabeth Benson ed., Dumbarton Oaks, Washington.

Etnia y sociedodr costa P eruana prehisponica. Institut0 de Estudios Peruanos, Lima

Rowe, Job H. 1946 Inca Culture at the Time of Spanish Conquest. In, Handbook of South Americm Indium Volume 2: The Andean Civilizations, ed. by Iulian H. Steward, pp 183-330. Cooper Square Publishers, New York.

1979 An Account of the Shrines of Ancient Cuzco. flowpa Pacha 17: 1-80.

Salazar?Ernesto 19ûS Investigaciones arqueologicas en Mdlumica (Provincia Del Pichincha). Miscelmiea Antropologica Ecuatoriana5:129-160. Salinas Loyola, Juan 1965 [1571] Relacion y description de la ciudad de Loxa In, RelacÏones geogrtijtcas de Indt'as9ed. by Marcos Jiménez de la EspeVol. 2 pp 29 1-306. Ediciones Atlas, Madrid. -

Salomos Frank Pochteca And MIndak A Cornparison Of Long-Distance Traders In Ecuador And Mesoamerica. Journui of the Steward Anthropologicai Society 9 (1-2): 23 1-247. Urbouia, Illinois.

Shamanism and Politics in Late-Colonial Ecuador. American Ethnologisî lO:413-28.

Native LorcLF of Quito in the Age of the Incas: The Poiitical Economy of Nonh Andean Chiefdom. Cambridge University, Cambridge.

Ancestor Cults and Resistance to the State in Arequipa, ca 1748-1 754. in, Resistance, Rebellion, and Cottsctot~messin the Andean Peu- World. 18th to 20th Centuries, ed- by Steve I. Stem. University of Wisconsin, Madison.

A North Andean Status Trader Complex Under Inka Rule. Ethnohistory 3% 1):63-77.

Salomon, Frank and George L. Urïoste 1991 7ne Huarochin' Mmuscn'pt: A Testanient of Ancient und Colonial Andean Religion. Transiated fiom the Quechua by Frank Salomon and George L. Unoste. University of Texas, Austin.

Simano, Joan 1%7 La Relacion Simano-Xerex. ( 1525) Las RelacÏones Primitiims de la Conquista del Peni. Raid Porcas Bamnchea, ed. pp 64-68. Lima.

Sandweiss, Daniel H. 1988 The Fishermen Of Chincha: Occupational Speciaiizatïon On The Late Prehispanic Andean Coast in, Economic Prehistory of the Centrai Andes, ed. by Elizabeth S. Wing and Jane C. Wheeler, pp 99-1 17. BAR International Sefies 427, Oxford.

1992 The Archoeology of Chincha Fishemen: Specialization and Stutus in Inka Penr. Bulletin 29 of Carnegie Museum of Natural History, Pittsburgh. Santo Thomas, Domingo de 1951 [156û] LéXicon, O va:ubtlImCrrtode la lengua general del Peru. Edicion Del Instituto De Historia, Lima

Schele, Linda and DaMd Freidel 1990 A Forest of Kings: ne (Intold Slos> of the Ancient Moya. Quill New York

Schortman, Edward M. 19119 Interregional Interaction In Prehistory: The Need For a New Perspective. American Antiquity, 54: 5265.

Schorûnan, Edward M. and Paeicia A. Urban 1987 Modeling Interregional Interaction in Prehistory. In. Advances in Archaeologicol Method and Theory Volume Il, ed. by Michael B. Schfler, pp 37-95. Academic, San Diego.

Shackleton, Judith and Henry Elderfieid 1990 Strontium isotope dating of the source of Neolithic European Spondylus shell artefacts. Antiquiîy 64:3 12-3 15.

Sharon, Douglas 1978 Wrfwdof the Four Winds: A Shan'sStory. Free Press, New York.

Shimada, izumi 1982 Horizontal Archipelago and Coast-Highland Interaction in North Pem: Archaeologicai Models. In, El Hombre y su Ambiente en los Andes Centrales, ed. by Luis Millones and Hiroyasu Tomoeda Senri Ethnological SuesNo. 10: 137-210.

1985 Perception, Procurement, and Management of Resources: Archaeologicai Perspective. in, Andean Ecology md Civilization, ed by Shozo Masuda, Izumi Shimada, and Craig Momk, pp 357-399. University of Tokyo Press, lapan.

1987 Horizontal and Vertical Dimensions of Prehistoric States in North Pem In, The Origins And Developrnent Of The Andean Skzte, eds. Jonathan Haas, Shelia Pozorski, and Thomas Pozorski, pp 130-144. Cambridge University, Cambridge. Cuiauaj Continuities and Discontinuities on the Northem North Coast of Penr, Middle-Late Horizons. In, The NorthDynasties: Kingship and StaiecrQt in Chimor, Michael E- Moseley and Alana Cordy-Collins eciitors, pp 297-392. Dumbarton Oaks, Washington.

Pachacamac Archaeology: Retrospect and Prospect. Introduction to Pachacamac: A reprinr Qthe 1903 edition by Mar Uhle, pp XVI- LXVm The University Museum of Archaeology and Anthropology, University of PennsyIvania, Philadelphia

Irene Marsha Mmn, Sun, And Devil: Inca And Colonial Transformations Of Andean Gender Relations. (Volumes 1 and II). University Microfilms International, Ann Arbor.

Silvennan. Heiaine 1988 Cahuachi: Non-Urban Cultural Complexity on the South Coast of Perm Journal of Field Archaeology 15:403-430.

1993 Cahuachi in the Ancient Worfd-University of Iowa, Iowa City.

Sola, Donald F. 1%7 Grmatka del Quechua de Humruco. Universidad Nacional Mayor De San Marcos, Lima

Stahl, Peter W. and Presley Norton 1987 Precolumbian Animal Domesticates From Salango, Ecuador. Antericon Antiquify 52:382-39 1.

Stiglich, Germb 1922 Diccionmio geogrofico del Perii. Torres Aguirre, Lima

Stone, Rebecca 1983 Possible Uses, Roles, and Meanuigs of Chavin-style Painted Textiles of South Coast Peru. In, Investigrnom of the Andean Part, ed. by Daniel Sandweiss, pp 5 1-74. Comell Latin Amencan Studies Program, Ithaca

Stothert, Karen E. 1993 Un sitio de Guangala Temprmo en el stvoeste del Edor. National Museum Of Naairal History, Washington; Museo Antropologico, Banco Central del Ecuador: Guayaquil. Tello, Jdio C. 1%7 Medio geografico andino. In, Paginas Selectas, pp 1-26. Universidad Nacionai Mayor De San Marcos, Lima

Tello, Julio C. and Tonbio Mejia Xesspe 1979 Paracas Segimcio Parte: Cavemas y NectOpoIis. Universidad Nacional Mayor De San Marcos, Lima

Thompson, Stith 1955 MoaiIndex OfFd-LLirerature: a class~ftcationof narrative elements in folktoles, bailads, myths. fables. rnediewl romances, exempla, fabliaux, jest-books and local legenk V.6 Index- Indiana University, Bloomington.

Topic, John R 1977 nte Lower Class At Chan Chan: d Qualitative Approach. PhD. Dissertation. Harvard, Cambridge.

Topic, Theresa Lange 1989 Ritual Offerings At Marcahuamachuco. Paper presented to the 2% Annual Meeting of the Institute of Andean Shidies January 6-7, Berkeley.

Towle, Margaret A. 1%1 nie Ethnobotany Of Pre-Columbian Pen. Aldine, Chicago.

Uhle, Max 1922 Influencias Mayas en el Alto Ecuador. Boletin De La Acudemia Naczonal De Historia N:205-240-

1991 [1903] Pachacamac: Report ofthe William Pepper M.D. LL-D.., Penivian Expedition of 18%. In, Pachacamac: A reprint of the 1903 edition by MiLlhle, Part U, The University Muswm of Archaeology and Anthropology, University of Pennsylvania, Philadelphia.

Verano, John W. and Alma Cordy-Collins 1986 HI MI :A Late Intemediate Period Mo- Structure at Pacatnamu. In, The Pacatnansu Pupers Volme I, ed. by Christopher B. Donnan and Guillenno A- Cock pp 85-94. University of California, Los Angeles. Villasante, Salazar de 1992 [1568/71] Relacibn general de las poblaciones espatioles del Peni. In, Reluciones hi~orico-geogrrificasde ia Audr'encia de Quito Sig10 XYI-MX ed. by Pilar Ponce Leiva pp 59-65 Consejo Superiot de Investigaciones Cientificas, Madrid.

Wallace, Dwight T. 1991a A Technicd and Iconographie Anaiysis of Carha Painted Textiles. in, Pmaars Art and Architecture: Object and Context in South Coastai Pen, ed. by Anne Paul, pp 6 1-109. University of Iowa, Iowa City.

1991b The Chincha Roads: Econornics and Symbolism. In, Ancient Roud Networks und Sedement Hierarchies in the New Worid, ed. by Charles D. Trombold, pp 253-263. Cambridge University, Cambridge.

Weigand, Phil C., Garman Harbottle and Edward V. Sayre 19T7 Turquoise Sources and Source Analysis: Mesoamerica and the Southwestem USA.. In, Exchange Systems in Prehisory, ed. by Timothy K. Earle and lonathon E- Ericson, pp 15-34. Academic, New York

West, Robert C. 1%1 Aboriginal Sea Navigation Between Middle And South America Arnerïcan Anthropoiogist63 :133- 135.

Whitecotton, Joseph W. and Richard A. Pages 1986 New Worid Precoiurnbian World Systems. In, Ripples in the Chichimec Seu: New Consideratiom of Southwestem- Mesoumericm Interactions, ed. by, Frances Joan Mathien and Randall H. McGuire, pp 183-204, Southem Illinois University, Carbondale.

Wilcox, David R 1986 A Historical Analysis of the Pmblem of Southwestern-Mesoarnerican Connections. In, Ripples in the Chichimec Sea: New Consideratiom of Southestem-Mesoamen'can Interactions, ed. by, Frances Joan Mathien and Randdl EL McGuire, pp 9-44, Southern Illinois University, Carbondale.

Willey, Gordon R 1972 The Artifacts of AI= De &criflcios. Peabody Museum, Cambridge. Wïlley, Gordon R and John M. Corbett 19% Eurij Anch And E* Svpe Culture: Chavin Horiron Sites of the Centrai Pe-an Coast. Columbia University Press?New York.

Wright, Henry and Melinda Zeder 1977 The Simulaiion of a Linear Exchange System Under Equilibnum Conditions. In, Exchange System in Prehr'story, ed. by Timothy K EarIe and bnathon E. Ericson, pp 233-253. Acadernic, New York.

Zeidler, James A. 1978 Pnmitive Exchange, Rehistoric Trade and the Problem ofa Mesoarnerican-South American Comection. Journu2 of the Steward Anthrop~io~calSociety 9: 7-39. Figure 1. El Segundo Mes, Febrero. Pavcar Varai Quilla Sacrificio con oro y plata y 10 rrecibe, y mollo y cuui. (Guaman Poma 1980 /238[24O]:2 12) Figure 2 Idolos 1 Vacas De Los Chinchai Svivs. Los yndios ... sacrificauan con pems porque ellos cornian perros y aci sacrificauan con el10 y con coca y comidas y sangre de perro y &o. (Guaman Poma 1980 /Xi6 [268] %û;/267[269] 241) Figure 3. tdolos I Vacas De Los Andi Svios. Sacnficauan Ande Suyo ai serro y uaca idolos de Saua Ciray, Pitu Ciray con dos nifios y conejos blancos Figure 4. tdolos 1 Vacas De Los Colla Svios. Sacrificauan con carneros negros y Sestos de coca y con dies niilos de un aiIo y conejos y mdoy pluma se suri. . .con cameros de cuyro y baxilla de barn> y mucha chicha de canaua y moraya y molIo, comidas y pescado fiesco y seco. (Guaman Poma 1980 /270[272] :244427 1127311245) Figure 5. fdolos 1 Vacas De Los Coude Svios. Sacrificauan la uaca idolo de Coropona, Urco, con on> y plata y con ninos de doze aiios y plumas de pariuana y de uachiua y coca y mu110 y srmco y came cruda y sangre cnda, con ella hazîendo sanco. (Guaman Poma 1980 /272[274]: 246;/273[275]:247) Figure 6. Paccha in the form of a plate of Spondylus from Chimu (C&h Cachot

- - - 4955.38) ------* princeps princeps (interior) showuig coloured margui.

princeps princeps (exterior) with scuiptured radials.

Figure 7. Spondylus princeps Figure 8. Spondylus princeps leucacaniha showing white coloured le@ spines.

Figure 9. Spondylus calcifer showing irregular shape and lack of prominent spines. Figure 10. Spond'ylus princeps unicolor. 148

princeps prhceps

princeps unicolor

Figure 1 1. Spine pattern differences on princeps unicolor and princeps princeps. Figure 12. Rollsut of the Tel10 obelisk at ChaMn showing Spondylus shell in the upper left corner and inset (Lyon 1978). Figure The Srniliag God firom Chavin de Huantir holding a Spondylus shell in the leftlfemale hand and a Strombus shell in the right/maie hand (Cordy-Collins 1978)- Figure 14. Staff god variant with "wingt'and pouc h of Spondylus shells. Chavin style motif present on Carhua textiles fiom Paracas area (Wallace 199 k75) Figure 15. Statf god variant with staff, headband and waistband finiais of Spondylus shells. Spin directions ofyarns show the specimen (C-4 ) was producd outside the south const vicinity of Paracas. (Wailace 199 W6) Figure 16. Inca cosmology fkom he Coricancha in Cuzco. Based on the drawing by Santacruz Pachacuti Yamqui (fiom Regina Eiamson l982:77). Figure 17. Staff Goddess from a Carhua-Chah textile illustrating the 'vagina dentata' . Figure 18. Staff Goddess fiom a Carhua textile showing the 'vagina dentata' (nom Bruhns 1994: 168). Figure 19. Spondylus shell diving scene taken from a Middle Sican (Classic Lambayeque) silver eanpool in the collection of the Museum ofthe Amencan Indian, New York (Norton and Marcos 198 1: 148) Figure 20. Middle Sican gold earspool with Spondylus diving scene hm Lambayeque. Peru (Bruhns 1994:284). Figure 21. Middle Sican goïd earspool Spondylus and conch shell diving scene (Cordy-Collins 1990:40).

Figure 22. Middle Sican metal earspool with Spondylus diving scene (Cordy-Coll ins 1990:401). Figure 23. Rollsut drawing from a Middle Sican silver cylinder showing Spondylus diving (Cordy-Collins 1990:404).

Figure 24. Textile scene with divkg motif fkom the North Coast of Peru with Spondylus shells represented (Cordy-Collins 19%:4O3). Figure 25. Middle Sican spatuia showing a Spondylus shell diver and the oblong weight attached to the back (Cordy-Collins L990:407). Figure 26. Location of Mullu Toponyms in Ecuador

1. Mu110 Corral 2 Mullo Pungu 2, Mullos 4, Mullos 5, Mulluleo 6. Mullunuma 7. Mullupungu 8. Mulluturo 8. Mollotoro 10. Mullumica Figure 27. Hatvn Chasqvi, ChwMvllv Chasqvi ,CmaChuni muilo chasqui [que trae caracol]. Guaman Poma 1980 /350[352]: 322;1351[353]:3323) Figure 28. Ecuadorian balsa raft with 'cabailitos' in the background Illustnited by Giralomo Benzoni in the mi& 1500s (hmBruhns 1994:365). Figure 29. Construction view of a sea-goïng balsa raft used by the Guayaquil coastai traders at the theof conquest ( Porras 1987: 137 [original by E. F. Paris, 184 1- 18451). OCEAN e\& e\&

Figure 30. Geographicai disûibution of Sp~~ushabitat and possible routes of Spndjius exchange system as suggested by Jorge G. Marcos (fiom Marcos lW7/78: 100). * QU^ - COLOMBIA -i. 21 7 i 28,: t ,1 i f '-\-. f 16: PERU ;-* E -/-

Figure 3 1.

14 ChanChan 28 LosTayos 15 Chavh de Huantar 29 Marcahuamachuco 16 Chinguilanchi 30 NepeiiaValley 17 EI ParaiÏ= 31 PampaGrande 18 Garagrry 32 Pikillacta 19 Ica 33 Pirincay 20 IngaPirca 34 Raya Miller 21 Islade la Rata 35 Pdremde Payogasta 22 Kuntur Wasi (off map) 23 LaGaigada 36 Puhape 24 Lalibertad 37 Rbplaya 25 LaPaioma 38 Saiango 26 LoDemas 39 SantaEiena 27 Los Frailes 40 Valdivia 41 Wari Kayan Figure 32. Physicai Appearance of Spondylus Materiai hmConfimieci Archaeoiogical Sites

Whole Shells ShelI framnents Worked Shell

Alacrh Island Aspero Atacames Batan Grande Cabeza de Vaca Cahuachi Caleta Vitor Cerro Blanco Cerro Copiapo Cerro ÈrafMique Cerro Narrio Cerro Vick Chan Chan Chavin de Hdtar Chinguilanchi El Paraiso Garagay ka inga Pirca isla de la Plata Kunîur Wasi La Galgada La Lihrtad La Paloma Lo Demb Los Frailes Los Tayos Marcahuamachuco Whole Shells Shell fbments Worked Shell

Morro de Etan Nepeaa Valley Pampa Grande Piruicay Playa Miller Potrero de Payogasîa Puémape Ricaplaya Salango Santa EIena Valdivia V-124 V-304 Warï Kayan Composite-head figurine with ducks. Common flat-head figurine. Chubby Flat face and polished on both sides. cheeks and forehead with mother-of- peul inlays in the eyes.

Figure 33. Fiprines fiom Cerro Narrio. (Banco Central Collection, Quito, Ecuador). Coast

Inca

Milagro - Quevedo Canari Cerro Nanio iII

Guangala Guayaquil

I Los Tayos Cerro Nanio 1

las Vegas

Figure 34. Chronology of Ecuador- Cultures and sites mentioned in the text or related to SpomtyIus activity (after Porras G. 1987). North Coast central Coast South Coast

Late Intermediat~ Period

Middle Horizon I

Moche Period

3riy Horizon Cupisnique Initiai Period

La Rorida Huaca Prieta El paras-so La Paioma

Lithic Period

Figure 35. Coastal chronology of Peru (der Moseley 1992). Time North

Late lntermediatc Period

Middle Horizon

Earfy Interrnediat~ Period

Initial Period

Preceramic Period

üthic Period

Figure 36. Chronology for highland Peru (after Moseley 1992). AppendixA

Sites Coataining SpondyIus Material -Site:

Location: About 500 metres off-shore of Arica, just south of Chfle/Pem border,

Period: Late Horizon

Contents:

Quantity of turquoise and thick red Spondyhs shell beads (Büd 1943: 19 1)

Comments: -Site: Ancon

Location: Ancon Valley, Pen Between the Chillon and Rimac Vaileys and the Chanmy Valley.

Period: Middle Horizon and Late Horizon

Contents:

Middle Horizon 2 burial inciudes press-molded; blackware face-neck vesse1 with body representing Spondylus. Middle Horizon 4 burials include same as above but with derivative Spondylusbodies (Menzel1977:46-47). . - Late Horizon burials from the sector known as Pampa de San Pedro or Barrio de las Esteras, include chaquiras and Spondylus valves (Ravines and Stothett 1976: 153-206).

Comments:

While the Middle Horizon 2 buriais do not include Spondylus material, they are included to emphasize the use of Spondylus as an object of importance that was displayed in various 0th- media-

The Middle Horizon finds should not be confused with the Early Ancon-Supe excavations reported on by Willey and Corbett (1954). This early work dated Ancon to the Formative or Developmental Stage, contemporaneous with Chavin and possibly related to the site of Kuntur Wasi. No Spondylus appears in the authors' Iists of identifid mollusks. -Site: hpem

Locations: Supe Valley, CO& Peru

Period : Preceramic

Contents:

Spondylus shell (Feldman 1992:73).

Comments:

Due to a lack of specific published documentation, the actual amounts, shapes. or sizes of the Spondylus fiom this site are difficult to determine. However, Spondylus fkom the site is referenced to Robert Feldman's unpublished dissertation by several other researchers. In any case the arnount of Spondylus in question appears to be quite smail, and possibly only a fraement.

See also La Palorna -Site: Atacames

Location: 25 km south of Esmeraldas, north coast of Ecuador

Period: Integration Period (Middle HonZon/Late Intermediate Period)

Contents:

Cuentas and chaquiras of Spondylus princeps and calcifer (Cabada 1989:98- 99; Gaivin Garciaand Bamuso Pérez 1986:6 1-63).

Materiai found in the 'chimney' in Tola E- 10 1 uicluded cerarnics, Strombus. Murex. SpondvZus. Pinctada rnazatfantica (Heras and Adanez Pavon 1989: 150).

An abundance of Spondylus princeps valves found in the coral reefs surrounding the Bay of Atacarnes (Guinea 1989: 139).

Comments:

The relationship between water and Spondylus may possibly be represented in Tola €401. While the chimney style is indicative of a type of burial an altemate explanation is provided by Heras and Adhez Pavon (L989: 150) for their use. None of the chimney tombs excavated held human buriais, but al1 held water (Hem and Adhez Pavon l989: 157). The suggestion is that the chimneys acted as water purifiers (Heras and Adiinez Pavon 1989: 159).

The inclusion of Spondylus in a mechanism intended for the purification of water (creating fresh water) would be expected in Peru, but the concept does not appear to have extended to Ecuador. The evidence is not overwhelming, as Spondylus does not appear in al1 the chimney tombs, nor was it found alone. Its association with Strombus and mother-of-pearl indicates that they al1 held value for one reason or another, whether they were part of a human buriai, or a ritual for Gesh Mer. -Site: AElngaro

Location: 15 km northwest of Huari in the Ayacucho Basin, Pem

Period: Middle Horizon

8 The period of occupation for this site is based on stratigraphy, midden accumulation and ceramic composition which suggest a 100 year maximum range fiom the eariy 9th to early lûth centuries. This period is confirmed by radiocarbon dates ofAD 760 +-75, AD 880 +-50,990 +65 (Anders l986:XO).

Contents:

Excavated ritual paraphernalia included: 25 ceramic vesseis, 19 ceramic figurines, 3 stone tablets, 1 ceramic tablet, 2 smail purnice cones, 9 turquoise seed eff~gies(corn, bean, pacae pod, gourd, 2 unidentified seeds), 12 geometric turquoise forms, 32 geornetric shell fonns, and 24 shell fragments (mostiy Spon&lus princeps spines).

Objects of adornment included: 17 copper women's clothing pins, 2 copper pendants, 66 pendants of a silver-copper alloy, 10 turquoise beads, 6 Spondylus princeps beads, and one bone bead (Anders 1986:2 1 1).

Azingaro is a Huari site. There are two types of building shapes, irregular and regular, at the site. Anders identified the irregular shaped buildings as residences of the more important, or preeminent, memben of the community. This conclusion was made on the evidence of heavier consumption of chicha, a sign of hospitality and generosity, and on the frequency of lwury goods found in them.

Items of rare and valued material, including turquoise and Spondylus princeps used for adornment and ritual, were found more frequently in these irregular buildings (83.4%) along with (76.2%) of the prestige pottery (Anders 19862 12). -Site: Banduras

Location: Moche Valley, north coast of Pem.

Period: Late intemediate Period / Eady Chimu

Contents:

Buriai EC 5, a young adult found with 4 ceramic vessels; 3 metal objects located in the chea region: and 1 Spondylus valve near to the metal objects (ûonnan and Mackey 1978:23O).

Comments: s The EC 5 burial was one of 8 burials purchased fiom huuqueros who had looted the cemetery at Banduras, and who were intewiewed about the site by researchers shortly afterwards. -Site: BathGrande

Location: Lambayeque Valley, Pem

Period: Late Intermediate Period / Middle Horizon Middle SichAD 850-900

Contents: r Sponddus was found in the Huaca Rodillona used as fil1 in column boxes, alternating with bundles of 1-shaped arsenical copper "foil", or naipes (Shimada Wû:341).

400 whole Spondylus shells were found buried with 200 human sacrifices and 2,000 bundles of arsenical copper foil (Shimada 19901367). r A number of Middle Sich misftom the site are made of gold and silver, and are decorated with representations of the Sich Lord or animais, with turquoise and Spondylus idays (Carcedo Muro and Shùnada 1985:69).

Tumi MOP-2708 has a blade and handle of silver, the head of the Sican lord with turquoise bead eyes, ears decorated with turquoise and gold turquoise earrings. The headdress consists of a cap with inset turquoise roundels and red Spondylus sheli (Carcedo Mun, and Shimada 1985:70).

Contents: r Tumi MOP-2444 has three decorative Spondylus shells on top. This is an indication of the importance of the sheil. The incorporation of representations of Spondyus into artwork aiso appears later in the cenunic art of the Chimu.

In the tombs and grave goods copper appears more commonly as the metal of the general populace while gold and silver is identified with the elite (Carcedo Mur0 and Shimada 1985:70). -Site: Cabezadt Vaca

Location: Tumbes, Peru On the left bank ofthe Tumbes river, about 6 km hmthe modem city.

Penod: IncaKate Horizon

Contents:

r A workshop site that included cornpiete Spondylus valves and anifacts in various stages of being cut or polished - both princeps and calcijk Also Anadara. Ostrea Strombus. Conus fergusoni and melongena parula (Hocquenghem and PeiTa Ruiz 1994211).

Comments:

r The Spondylus remaùis at Cabeza de Vaca are descnias follows:

" De Spondylus princeps y Spondylus calcifer se encontraro n conchas completas, valvas completas a veces totalmente pulidas, valvas cortadas desde el umbo, partes laterales completas, partes cortadas y pulidas, charnelas. Se recuperaron también trozos cortados con desgaje, cortados sin pulir? pedazos de posibles concreciones sin acabado, de posibles nodulos, nodulos. Finaimente se recogieron contomos y siluetas de figuras, figuras sin acabado, figuras y fragmentos de figuras y finalmente cuentas y petos" (Hocquenghem and Peila Ruiz 1994214).

Cabeza de Vaca is viewed by Hocquenghem (1 994:224) as an administrative centre that acted as a terminal for the transport of exotic shells hmthe north to the central Andes. The shells would arrive by balsa raft and then be transported inland by llama caravan. The shell would be worked dong the route at tmbos such as Rica Playa (see Appendk A - Rica Playa), a one day walk hmTumbes.

Also known in the literature as Cabeza de Huaca, San Pedro de los Incas and Corraies or Tumbes Viejo. -Site: Cahuachi

Location: Nazca Valley, Peru

Period: Early Intennediate Period Nazca 8 (appmx. AD750-850)

Contents:

- At the base of Unit 19, in the Room of the Posts, ten complete, unworked Spondylus shells found interred in the sand filling a niche with two &ers found in a circula depression in front of it (Rodrigue2 de Sandweiss l993:294; Siiveman 1988:417; I993:178).

From the surface of the upper part of the Unit 19 mound a trapezoidal fiagrnent of Spondylus was recovered (Silveman l993:US).

On the north slope Unit EE, a flat-topped hill, several large pieces of Spondylus shell and a few decorated Nasca sherds were found (Silveman 1 993 :8û).

Comments:

Only eleven shell artifacts were recovered at Cahuachi, one of which was the Spondylus trapezoidal fragment fkom the surface of Unit 19 (Silvennan 1993 275). -Site: Cakta Vitor

Location: 30 km south of Arica, Chile.

Period: Colonial

Contents:

Mummy bundle hm1578 Uicludes a matrire male with colonial period artifacts and clothing and a short string of Spondylus beads as well as native and colonial arti-fa(Bird 1943 :250). .

Comments:

Dated by the inclusion of a paper, the Proclamation of Indulgences, folded against the body (Bird 1943 1250)

Indicates both the extent of Spondyhs movement and its continued use in the post-conquest era. -Site:

Location: Central CordrlIera, Chilé

-Period: Late Horizon

Contents:

Red and white Spondylus figu~esshaped Iike Ilamas and humans (kiÏ.amm Charlin 1978:447; Reinhard 1%Mû).

Comments:

0 These tlgurines are known as iIlas or enqas used as power objects in sacrificial contem.

Possibly found in this location as a result of Inca militaq expansion. -Site: CernRaîiaiiique

Location: Chulucanas, Pima Valley, Pent

Initial PerÏod

Contents:

Spondylus working debris (Burger 1992:10 I ).

Comments:

A hill site overlooking the town of Chulucanas that was reworked into a series of ascending stone-faced platfonns connected by inset staircases. By the Early Horizon the architecture began to resemble a U-shaped layout (Burger 1992: 100). Possibly a gateway commun@ for goods moving through the PimValley across the Andes (Burger 1992: 10 1). -Site: Cerro Natrio ( includùig Chuquipata, Chordeleg and PindiIig)

Location: Caiiar, Ecuador

Period: Formative (PreceraimidInitial Period/Early Horizon)

Contents:

Spondylus figurines, cuentas, pendants, ear spools, fragments, found in îhe lower levels at Narrio (Collier and Murra 1943).

Comrnents:

Possibly a central point of trade and workshop activity on an east-west route across the Andes,

The tigurines (Figure 32) are unique to Cerro Namo, but other matenais and Spondylus artifacts may have been worked and passed on through the tracie route. Spondylus is also found east of Cuenca at Los Tayos, on the eastem slopes of the Andes.

The Spondylus figurines nom Cern Narrio and smunding area are presently located in many private and public collections. Provenience is poor for actual site ongin, leaving most of the collections fiom the Cailar Valley (including Verneau and Rivet's finds fkom Chuquipata, Chordeleg and Pindilig) generically described as Narrio.

There were two distinct periods of activity at Cerro Narrio. The Spondylus finds are part of the Eariy complex. Spondylus material does not appear among the Late complex cultural traits as listed by Collier and Murra (1 943 :8 1- aI). -Site: Cerro Vicus

Location: Upper Piura Valley Peru.

Period: Initiai period

Contents:

Spondylus debris (Richardson, McConaughy, Heaps de PeÏia, and Décima Zamecnik 1 99O:438).

Comments:

- Possibly related to Cerro Raiiailique (see above). -Site: ChanChan

Location: Moche Valley, Pem

Period: Late Intermediate Period

Contents:

Spondvlus found in sub-floor burials in the audiencias (Keatinge l982:203).

Six complete Spondylus shells in an olla under a kitchen floor in the SIAR (J. Topic 197T84).

Spondvlus is found in the royal burial platforms, used whole, ground or crushed into powder [Conrad l982:99,lO4).

Also a stone-lhed bin of ground Spondylus was found in the forecourt of a buriai platfonn (Conrad l982:%),

In the Huaca El Dragcin the central shrïne offerings included Spondylus shell and worked shell fragments [Menzel1977:41).

An area at the north end of a bench running along the West wall of the forecourt of the burial platform in the Ciudadela Rivero was cleared. Beneath the clay plaster a L/2 cm thick layer ofground Spondylus shell was discovered (Day 1973 :212).

Loose earth of the backdirt pile hmthe southwest corner of the burial platform contained over a dozen Spondylus valves and six Conus shells. Both shells were found in profusion in the looters backdirt of the burial platform in Ciudadela Tschudi (Dav 1973:2 1 7 Francisco Iriarte 1969'1 b Commentsr

"Spondylus and Conus shells in the backdirt or burial platform, along with ground Spondvlus sealed in the floor of a bench in the forecourt are indicative of speciai status or ceremonialism"(Day L 973 D6).

+ Spondvlus beads are quite common grave goods from the tombs at Chan Chan. Because of heavy looting the context is poor, with no way of knowing whether diey are fiom iow, middle, or high status burials. Many of the multi- media artifacts found looted and in context at the site have inlatd Spondyus (Theresa Topic 1995: personai communication). -Site: CbavUi deHuaintar

Location: Ancash. Pem.

Period: Late inïtiaEarly Horizon

Contents:

Initial Period:

From the Old Temple, the main passageway contained a femaie hurnan sML about 30 - 40 years of age, surrounded by a circie of40 rniik teeth. On the floor were ceramics with the bones ofcamelids, deer, guinea pigs, fish, musse1 shells and cut fragments of Spodjdus (Burger 1992: 138).

Early Horizon:

Two lots of Spondylus shell with remains of guinea pig were found in dedicatory offerings interred in the retaining wall and floor of the Janabarriu platform. 53 fragments were found in total, including 12 beads, 6 cut pieces and 35 fragments that may have been refuse. It appears that al1 the pieces were of princeps and not calcifer(E3urger 1984:214JS7,1992: l69,l7 1).

Comments:

Chavin is an important Spondylus site not only due to its physical appearance in the archaeological record, but also because of its visual representation in the Chavin iconography .

The conceptual traits associated with mullu of SpondylusAefVfemaie and Strombus/right/male are derived ftom representations of the Smiling God dated to Phase D at Chavin, -Site: Chiicba

(including: La Centinela, Huaca de Alvarado and Lo Demis)

Location: The Valley of Chincha, Peru. Approximately trïangular in shape, extendhg some 10 kilornetres in length and 15 in width.

Period: Late Horizon

Contents:

From Lo Demhs a single Spondylus ment(Daniel Sandweiss 1988: 102).

Burial E-13 frorn Site E fkom Pampa de Canelos in front of the mins of La Centinela in the Valley of Chincha, about a kilometre inland, included strings of mal1 red Spondylus beads and green glass beads (E 13-3975). Showing evidence of being made derthe conquest (Kroeber and Strong 1965:g).

One possible LCa II (Late Chincha transitional to Inca) burial, E6, contained two small Spondylus shells. Inca graves E 12, 13, 15 contained fiom 2 to 5 Spondvlus shells each (Kroeber and Strong 1965:30).

Site D. referred to by Uhle as Proto-Chincha, was a collection of previously opened chamber-like tombs in Huaca de Alvarado, a mound-like huaca fiom the last pre-Inca period (Kroeber and Strong L 96551 ).

Artifacts were assigned to LCa I (Late Chincha - early) or Inca. To the Inca were assigned fine beads and Spondylus pendants (3 850,385 1,3852). Uhle felt that the mound structure was older than the tornbs.

From this earlier penod he found several artifacts associated with skulls different fiom those of the LCa II and Inca burials. Including: a bead of lapis lazuli (3892~)and small worked pieces of Spondylus shell(3892d) (Kroeber and Strong 1%Hz). . Comments: a Uhle (Kroeber and Strong 1%5:3 1) concluded that Spondylus and fine beads were characteristic of the inca buriaIs, but vimaily lacking fiom both LCa I and II burials. Because of this he assumed that trade in Spondyius did not reach significant proportions until the Inca penod. r ühle's description of the necklaces, pendants and beads is found in the folloning with an expianation of the green glas beads that were of European origin:

" In beads and necklaces, the LCa and Inca graves contrast as markedly as in their Spondylus shell content. The characteristic LCa II ornament would appear to be represented by the necklace C 3-3725. A number of white shell omaments are in the shape of animals or birds. More are quadrilateral and bored with several holes which were inlaid with green Stone. Several of the pendants still retain one or more of the insets.

By contrast, the Inca graves contain pendants of Spondylus which are regularly rounded oblong, never in the form of animals, and without inlay. Such are E 13-3980-398l.3983,3984,4ûO6; E 12-396%

Also characteristic of the Inca graves are fine evenly-rolled beads of pink Spondrlus, white shell, violet shell, and black material. In some cases these average barely 2 mm. in diameter and i mm. in thickness. The finest beads corne in short loops, several of which are grouped together. On each loop the beads sometimes form a pattern of three colous. Such beads may have served as ear ornaments. E 13-3975 is an example of extreme fineness. E 13- 3975b has the individual Spondylus beads almost equally fine. The ends of the strings are tipped with green glas beads. these are larger, coarser, and far less evenly regular than the Spondylus beads of native manufacture. E 7, 1 1, 12,14- 15, and 18 contained loops of small beads sirnilar to those in E 13. The black beads preponderate, but pink and white ones are introduced.

As compared with the high development of the bead-making art in the Inca period? the LCa 1 and XI graves are entirely without srnall and regularly made beads except for the E 6-3943c and E IO-3957a On the basis of their pottery - one vesse1 each - these two graves were tentativety assigneci respectively to LCa II and LCa 1, The bead work would tend to throw some doubt on this finding. Funher, the E 10 string of beads was associated with three Iarge triangular Spondvhs pendants, and two cornpiete Spondylur sheils (3957e-f) were found in E 6-

E 33930d is an assemblage of very srnaIl beads of various sues and materials from a hitheno undoubted LCa I grave. There is a Spondylus pendant cut into step shape. The beads include three or four biue or green ones that look as though they might be glass. Too rnuch importance need not ,however, be attached to this occurrence of a European material in a LCa I grave because the beads in this lot were found ioose, as compared with those nom pure Inca graves, which reguiarly remain on the orig'inal string. There might be considerable chance for small beads to roll down into a grave, especially in excavation in sandy soil.

C3-3743,hmaLCaII grave, is a mass of beads of shell, Stone, pottery, and metal which, when re-stning, assemble to about enough length to enclose a neck. They are of a great varïety of size and shape, as weli as material. These beads may be characteristic of Chincha bead work previous to the pure Inca influence" (Kroeber and Strong 1965:30-3 I ). -Site:

Location: To the no& of Loja, Ecuador, on the bank of the Saragum river.

Period : Regionai Development (Eariy Horizon/Early Intermediate Period) Contents: a Offerings filled with green, red, and white muilus of stone and shell. Below the al= was a group of 40 Spondylus shells(Uh1e 1922:208).

Comments:

Dated to the Chauilabamba Phase that followed Cerro Namo and preceded the Caiiari.

The find is described by ühie as follows:

"Todo el subsuelo del piso contenia numerosas conchas de Spondylus (pictorum) votivas rellenadas con mullus verdes, colorados, blancos, de piecira y concha, y objectos de otras clases. En varias partes del subsuelo se encontraron también grupos de las diferentes clases de mullus, en estado libre; debajo del altar mismo, un gmpo de, rnk O menos? 40 conchas de Spondylus amontonadas" (Uhle 19U:2O8). -Site: Chiquitoy Viejo

Location: Chicarna Valley, no* cos* Pem.

Period: Late Horizon

Contents:

Spondylus fragments and valves found in both burial and administrative contexts, dong widi fineware ceramics and fancy textiles (Davidson 1980 14- 16; [Conrad 1974: 1641)

Comments: s This site was connected to the Vini and Moche Valley sites through its suggested use as an Inca administrative centre during the Chimu-Inca phase (Davidson 1 980 14- 16; [Conrad 1974: 1641). -Site:

Location: Vini Valley, north coast, Peru.

-Period: Early Intermediate Period

Contents:

Ear plugs inlaid with Spondylus and stone mosaic, with fmntals that range in size from about 8- 10 cm in diameter (Jones 1 97957). r One pair of ear plugs (Jones 197958 Figure 41) has a depiction of a pair of a bird-headed waniors carrying a mace, circuiar shield, and a sling with a golden projectile. The diameter is 102 cm with a length of 9.5 cm. This pair of ornaments were repaired in New York by Junius B. Bird and Charles Tournelli at the American Museum of Natural Histoty. Bird noted that originally there were mail pellet rattles in the shafts.

Another pair of lizard decorated ear plugs were found in same adobe-lined grave on a difXerent body.

The ear plugs were dated to the 4th cenw (Mochica III) in Vini based on provenienœ information nom Junius B. Bird.

Among the materials used in the mosaics are turquoise, chrysocolla, quartz, pyrite. and Spo~lusshell. Location: Rïmac Valley, Peru

Period: Initiai Period

Contents:

A dlSpondylus bead embedded in a votive offering (Burger 1992:63).

The votive offering as descnied by Burger ( 1992:63) was, "a small piece of granodiorite wrapped in Cotton thread, plastered with gray clay, slipped with shimrnering specular hematite, and then painted with the image of a supernaturd being with huge upper canines, and eyes with eccenmc pupils. Cactus spines were tied to its sides, as if representing staffs. Embedded in it was a small Spondylus shell bead brought fkom Ecuador. The investigators [Ravines and Isbell 19761 have linked this piece to the main supernatual image of the Old Temple at Chavin de Huantar. .." Location: Chancay, Peru

Period: Post Conques

Contents: s Spondylus shells fastened shut with small greenstone id01 inside and three small Stones representing the fim lima bean, wheat, and chili pepper (Cook l992:3S9,palcarWll933:51)

s The description of a srnall stone seed representing wheat indicates that this was a Colonial arrangement. However, the interpretation may have been made long after it was origïnally created, or perhaps represents a transitional phase. -Site: Ia

Location: Ica Valley, Pem

Late Horizon

Contents:

8 A youth surrounding the principal dead in a bufial was fiund with necklaces of Spondylus shell pendants, copper bell, pottery ocarina, and entire Spondylus shelt halves. Also wedge-shaped Spondyhs pendants (Menzel i977: 12J3)-

Comments: -SÏte: Inga Pima

Location: Caiiar, Ecuador

Contents:

Seven Spondylus fragments found entombed with 17 fernale skeletons in the Temple of the Moon (personal observation).

Comments:

The Temple of the Moon is fond in the eastem section of the site where structures cm be dated by their architectural construction and materials to the Cafiari culture phase. Western sections of the site, including the Temple of the Sun, were consûucted using Inca techniques after the Inca expansion.

Of the seven fiagments one appears to be a hinge piece with the spines ground dom; one is rectangular; two are rim fiagments; and three are triangular shaped pieces shilar to the shape of the figurines at Cerro Narrio (personal observation). -Site: Ida dt la Plata

Location: Manabi Coast- Ecuador

Period: Regional Development - Integraton (Eariy Horizon - Late Horizon)

Contents:

Spondylus princeps and culcifer dated to Guangala phase, Mantefio phase, and in conjunction with Inca artifacts. Excavations in the Bahia level included beads of Spondylus and pendants of green stone, some of them turquoise (Dorsey 190 1; Marcos and Norton 198 1 ;Norton 1986; Stahl and Norton 1987).

La Plata is important not only for the Spondylus material located there, but also because of the possible Inca burials identified by Dorsey (1901) that might be the remains of mullu chqui camaytoc, and therefore related to trade activity . -Site: Kuntur Wasi

Location: On the hiIl cailed La Copa, Cajamarca, northem highiands Peru

Period: Initiai Period/EarIy Horizon

Contents: r Fernale burial with 849 beads of red Spondylis in van-ous shapes and sizes. Unspecified number of smaller beads of Sponcfyius (Kato 1993 222).

Comrnents: r Five phases are established for Kuntur Wasi the most important being Kuntur Wasi with C-14 dates on ceramics of 760 +- 80 BC, 6 10 +- 60 BC 570 +- 60 BC 560 and 460 +- 50 BC (Kat0 1993:2OS). r Tombs excavated fiom the Kuntur Wasi phase hcluded several males with no Spondylus artifacts, but a male 50-60 years old was buried with Sbombus astifacts: 3 trumpets, 2 small Stronbus beads, 1 S~ombusplate, and I pendant.. A ritual relationship is suggested with the construction of the ceremonial building (Kato 1993:2 1 7)

+ It is important to note that the Spondyius is in the female tomb and the Strombus in the male tomb. There is a similarity between this gender association and thaî of Chavin de Hubtar. -Site: LaCabrya(G-l IO)

Location: Valdivia Valley, Manabi Coast, Ecuador

Period: Formative (Initial Perid)

MachaIilla Phase BC 1500-BC 1050

Contents:

Excavation of G- 110 included Spondvlus abraders and polishers (Meggers, Evans and Estrada 196%1 L 3). Seven of these artifacts were recovered but identification is restricted to either Spondylus or Sh.ombus. Also some wfinished pendant blanks of Spndylus.

Comments:

Worked shell objects at Machalilla Phase sites are of the same type as those found in earlier Valdivia phase A, B' and C sites. Pendants, beads and other objects of personal adornment are very rare (Meggers, Evans and Estrada l%S:l13) -Site: LaGalgda

Location: Santa River, department of Ancash, Pem. About 74 km hmthe Pacific-

Period:

Contents:

Female burial in C-10:ElO with necklaces having a centrai pendant of Spondylus and another skeleton with a necklace having a rectangular bead of Spondylus drilled laterally twice.

Initiai Period:

Cache with princeps shell disk carved with four bird heads (Grieder 1988:89,92,94; also see Burger 1992:S).

Also a Spondylus princeps shell disk carved with four profile monster heads in a Chavin-related style (Grieder l988:W).

Comments: s nie Spondylus use at the end of the Preceramic appears as personal jewellry dong with Strombus and 0th- shells, and turquoise (Grieder 1988187).

The Initial Period cache appears to have been an offering used to sanctiQ a ceremonid building (Grieder 1988:92).

La Galgada is situateci in a canyon halfway between the Pacific Ocean and the Amazon Valley. Radiocarbon dates for the late Preceramic indicate lwcury goods were traded with the Pacific Coast and the Amazon Basin. Inca ruins indicate its possible use as a route between the desert and tropical forest during the Late Horizon (Grieder, Bueno Mendoza, Smith and Maluia: 1988). -Site:

Locations: Santa Elena Peninsuia, Ecuador.

Period: Regionai Development - Integratïon Guangala-Engoroy-Mankfio Early Intermediate Pend - Late IntediatePeriod.

Contents:

Guangala: r Spondylus shell amulet and barre1 shaped cylindrical and discoidal beads (Bushnell1951 :62)

Engoroy :

At cemetery in La Libertad, following the Guangala period, a bwial of a child was found with two beads beneath the figure.. One a flat green stone and the other biconical ,together with an amulet of pink and white Spondylus in the fonn of a conventionalised human figure (Bushnell 1951 :87).

A large number of cylindrical Spondylus and copper beads apparently stning together with no definite system of alteration (Bushnell 1951 :62,87,99).

Comments:

Bushnell found the material difficult to distinguish fiom the coloured travertine deposited by the hot springs near Cuenca in Azuay. Beads were made fiom the buff-coloured travertine which is formed from calcium carbonate similar to Spondylus. -S ite: LaPaloma

Location: Central Coast, Pem.

Period: Early Receramic

Contents:

.r Small undetermineci fragment of Spondyfus(Fe1drnan 1992:73).

Comments:

+ Like Aspero, the Spondylus referred to for La Paloma is not well published, and as such Iittle is knom about the size, quantity, or shape of the sample.

See also Aspero. -Site: Los Frailes

Location: Machaiilla, CO& Ecuador

Penod: ïntegration Pend (Middle HorizonRate Intermediate Period)

Contents:

Rectangular Spondylus plaques, cafczYerplurnb bob and cuenias made of the red Spondvlus rim (Mester 1990).

C omments:

Los Frailes is more important for its possible mother-of-pearl workshop than SpondvIus. Close to Salango and Ida de la Plata, Mester argues that madreperla was also referred to as mullu in the chronicles. -Site: Los Gaviiancs

Location: Huarmey Valley, Department of Ancash on the coast of Pem

Period: PrecerarnicAnÏtial Period

Contents:

Frorn an excavation in Quadrangle D7 a clrenra of Spondylus princeps: 2.1 cm square with a large oval through the centre of 0.5 cm and perforated iongitudinally in two spots where string could go through (Bonavia 1982: 143).

Comments: r There were only four shell cultural elements found on the site by Bonavia, which included the Spondylus dated to the second epoch. Radiocarbon dates on maize in the pit indicate 1750 BC (Bonavia l982:74). -Site:

Location: Onente, Ecuador

Period: Eariy Formative (Preceramic)

Contents:

Four cornpiete valves of Spondylw princeps and worked pieces in the shape of trapezoids, perforated discs, cylindncal cuentas, a feline mask, a ring, and a pendant of a serpent-bird.

Located in the eastem foothills of the Andes and the Amazon basin, Los Tayos may have been a terminal in a trade route fiom the Amazon through Cerro Narrio and on to the Pacific Coast,

- The feline mask and serpent-bird characterizations predate Chavin iconography with similar representations. However, atDibuting feline characteristics to the mask appears to have been a very subjective judgment (personal observation). -Site: Mjircahuamacbuco

Location: Condebamba Valley, Pm.

Period: Early Intemediate Period/Middle Horizon

Contents: r 9.6 kg of Spondylus shell recovered Corn a pit in the Castille. Including complete half shells. cut rec*ui@ar pieces with drilled holes in the centre top and bottom, and a quantity of cut and carved bluish-green stone (T. Topic l989:3).

Comments:

Some of the bluish-green stone had been carved to look like srnall Spondylus shells (T.Topic 198913). This is sirnilar to the Moche and later Sich practice on the Coast OC making representations of Spondylus in various visual media

While people at Pikillacui were making figurines out of turquoise, people fiom Marcahuarnachuco were making Spondylus shells out of turquoise, and other Huari sites were including it in their buriais. The fact that turquoise was also carved into the shape of another ritually important object such as Spondylus is an indication of the importance of both materials. -Site: tWhe(Cerro Blanco, Huaca el Dragon, Huaca de la Luna)

Location: Moche Valley, Peru.

Period: Early Intermediate - Middle Horizon - Late Intermediate

Contents:

Spondylus shell and worked shell hgments dong with Conus ferpont in different stages of work towards becoming ornaments or inlays. Also destroyed shells, and some that had been incinerated (Bourget 1995; Menzel lW:41).

Comments:

The Uhle collection from Tmjillo includes 3460 museum entries. Of that catalogue about 1900 of them are items of whole or cut Conus and Spondylus (Kroeber L%5: 193).

The MocheKerro Blanco collection at the Phoebe Hearst Museum at Berkeley contains dozens of carvedkhaped zoomorphic and geometric Spondylus plaques (Theresa Topic 1995: personal communication). -Site: ~MomdeEten

Location: Lambayeque Valley, Peru.

Period: Eariy Horizon (Late Formative, E3C 500- BC 200)

Contents:

Ln the Temple of Mono de Eten there are layers of ground Spondylur that appear to have been used for offerings (Elera 1993 :2S2).

Comments: -Site: Pacatiirunu

Location: At the mouth of the Jequetepeque Valley, Pem

Period: Middle HorizonLate Entemediate ferîod

Contents:

Spondvlus shells, some charred, found in association with copper, textiles, gourds. seeds and Conus in burials of 8 young individuals between ages of 1 1 and 20 (Bourget 19954).

Four identifiable bunais were recovered from a mound Iocated in the soudieast corner of the Major QuadrangIe of the~uaca1 Cornplex.

Bucial I is a textile bundle contaking the body of a female approximately 20-25 yem old. The body lay on its back with the legs flexed and the knees drawn up on the left side. nie arms were crossed over the chest, and a valve of Spondylus princeps was tied to the paim of each hand with a sheer cloth band. The nwo valves of Spon&lus articulated, indicating that they were a single animal.

Buriai 2 was a female L 5- 18 years of age associated with elaborate textiles but with no Spondylus.

Burial 3 and 4 were infants, also without Spondylus, associated with low- status burials due to the quality of grave goods (Verano and Cordy-Collins l986:8%9O).

From the Audiencia room in the southwest Quadrangle three burial chamben were excavated by Susan Bruce (1986:97-100). While no specific date for the burial chambers has been identified, the bricks used in the architecture are associateci with Chimu occupation (AD 1100-1400) in the early part of the Late Intermediate Pend(Bruce 1986: 105). Buriai Chamber I is located in the centre of the u-shaped audiencia Human bones tound there represent one female 11- 13 years old and at least four other individu& I5-20 years old of undetermineci sex- The bodies were wrapped in textiles at thne of interment A large Conus fergusni shell with a textile band around it was associated with the skeletal material- Mixed with the human bones were whole and broken bricks, broken and some chaned Spondylus shells, ceramic sherds, lIama and small animal bones, plant remains, textile fragments, string, gourd fragments, wood spindle fragments, a copper spindle whorl, s small piece ofcopper shee~perforateci espingo seeds, and shells.

Burial chamber 2 is approximately 4.5 rnetres in front of the audiencia, and included at least one individuai 12 -14 years old of undetennined sex. The chamber contained broken gourd bowls and a mal1 gourd cinim. Sherds were identified as representing Moche? Lambayeque and Chimu. Also, fragments of copper omarnents made fiom hammered sheets, four Spondylus beads, and a carved wooden finiai,

Burial Chamber 3 included two sets of bones LZ-14 years of undetermined sex. Boner cluster 1 included a left hand holding a Strombus shell (see below). No Spondylus is rnentioned for this burial.

Comments:

Pacatnamu is an architectural complex of approximately one square kilometre. with more than 50 truncated pyrarnids. ïhere were severai different cultural occupations including Moche ca AD 600 - 900 AD, Chimu begiming ca. 1 100- 1 150 AD until about 1370 AD with the possible takeover of the Valley by Chan Chan, 150 kilometres to the south. It was a centre for ceremonial activity and political power for the lower Jequetepeque Valley.

The shell tied to each hand in Burial 1 (Verano and Cordy-Collins 1986:87-89) is interesting for the possibility that the tying of hands with Spondylus may be similar, or related to, the Paracas burial were the hands were tied with chaquiras of shell. Unfomnately there is no indication Ui the report from Buriai 1 of which side of the shell (left/righ&maielfemale) was in which hand. It would be interesting to know if the concepts of the IeWfemde vaive were matched with the lefi- hand symbolisrn that appeared on the Srniling God at Chavin de Huantar, or not.

If the leNright, maielfemale symbolism was known and in use at Pacatnamu then the left hand of the skeieton in Buriai Chamber 3 should have been holding a Spondylus shell, not Strombux This is the opposite fiom that of Chavin. but perhaps it can be seen as an indication that the female metaphor was not in use at this time or in this region. s It should also be noted that no turquoise is reported for this site. Instead the Spondylur is found in association with Conus fergusuni as at other north Coast sites such as Moche- -Site: Pachacamac

Location: On the Pacific coast of the Lurin Valley, about 18 miles south of LünaPeru

Period: Early Intemediate Period through to Late Horizon

Contents:

The door to the Pachacamac Temple, built of wooden sticks and covered on each side with cotton cloths, has one side showing a complex scene depicting Strombus surrounded by marine creatures, birds and plants, while the other side is plain and adomed with Spondylus (Pérez Bonay I975:32 1; Shimada 1991:XXIC[V)XXIC[V)

Excavations by Max ühle (1903:37) included two burials in the second stratum of the new soi1 in front of the Temple at Pachacamac. Burial A included a ciosed Spondylus shell (identified as pictorwn), and another half shell of the same species, and a poiished shell bored for suspension (photo indicates princeps). Burial B included a shell of Spondy1u.s pictorum

Neither of the graves is identified as to gender and are given as an example of the types of articles found in the mummy bundles which included articles of clothing, adomment and food. However, Uhle ( 199 1:39) identified these burials as containing members of the Inca nobility based on hair length and headdress.

Comments:

Ornaments found with female buriais in the Cemetery of Sacrificed Women included necklaces, bracelets and small pendants of shell that are reminiscent of chaquiru. One necklace made of several rows of shell includes carved animal fonns inlaid with mother-of-pearl. Also on the chah is a fkog carved from white shell with one green and one red eye. Some mnkets are embellished with beads or small pendants of blue stone identified at the time as either lapis Ianrïi or sodalith (Uhle 199 1:95). -Site: PampaGrande

Location: Lambayeque Valley, Peru

Pend: Middle Horizon

Contents:

Spondylus workshop in context with a copper working area adjacent to the Huaca II, with radiocarbon dates of AD 650 Thirty-two whole shells and hundreds of broken spines, chips and rough rectangular pieces (Shimada 1987: 137).

Also fiom Huaca Fortaleza, Spondylus shells, pendants and necklaces made fiom napezoidal Spondylus fkagments. Some found in association with llama bones and a child's skeleton (Haas l985:397,4W 1,404).

Comments: r A Moche V ceremonid centre (Shimada 1982:1 55). -Site: Paracas

(includuig: Arena Blanca, Necropolis of Wari Kavan and Cavernas, the cemeteries ofCerro Colorado)

Location: Paracas Peninsula, Peru

BC 6ûû- BC 175, Early Horizon

Contents:

Mummy No. 253 hmWari Kayan was found with a small Spondylus necklace and the fingers tied together with two chaquiras of Spondylus. Mummy bundle No. 1 14 Uicluded two necklaces of Spondylus (Tel10 and Meji Xesspe 1979).

Descriptions of mummy bundles numbered 290,3 19, and 383, each indicate the inclusion of a necklace of shell beads (Camion Cachot 1949124). Carrion Cachot ( 194959) identified these shells as Spondylus, used in both the necklaces and bracelets recovered fkom other mwnmy bundles.

The contents of bundle 3 10 fiom the Necropolis includes a Spondylur sheli necklace (Paul 1990:39).

Comments:

A comparison by Anne Paul (1990:42) of two bundles indicates that Spondylus appears in the bundle of an important person. Bundle 3 IO had more garments in total and more embroidered garments in conjunction with the Spondylus than bundle 378. While 378 had no Spondylus it had instead a necklace of bird-bone tubes. The ritual paraphernalia in the two bundles also indicates that 378 was of lower status or rank than 3 10.

Paul assumes, based on Paulsen (1974), that Spondylus was a symbol of higher ceremoniai status. -Site: Pikillacta

Location: Lucre Basin, lower Cuzco Valley, Peru

Middle Horizon

Contents:

Whole valves of Spondylus, worked pieces (some with intentionai perforations). and 5 worked rectangular fragments. In association with two sets of 40 turquoise-coloured stone figurines dressed in hats and tunics (Cook 1992: 344). The offering was buried below a floor, under a stone together with two valves of Spondylw (Memel 1968:s 1)

Comments:

Also found in context with Strombus and a copper (bronze) bar. The offering of figurines is considered to be in the H& style (Memel 1968:s 1). -Site: Pinilla

Location: Near Ocucaje, Ica Valley, Pem

Period: Middle Horizon

Contents:

A tomb consisting of two underground adobe-walled chambers, one on top of the other. niere was no burial in the upper chamber but it did contain Spondyhs fragments, some drilleci for suspension, and some cerarnic objects (Paulsen 1968: 1).

The lower tomb contained two skeletons in flexed position. The skeletons were designated Burial 1 and Buriai 2. Burial 2 included fragments of unworked Spondylus and 9 "bifid" Spondylus objects, each drilled with two holes (Paulsen 1968: 12).

Comments:

The description of Burial 2 follows: " Southwest of Burial 2, and in the southeast corner of the chamber, lay 10 pottery vessels gouped together, denoted in the field notes as Cache A. The region above the head of the burial produced 18 small tnunpet or lily-shaped cones of copper. the stem of each of which was twisted through a perforation in a srnail gold disc. Near the neck were fragments of unworked Spondylus shell, as well as 9 "bifid objects carved from Spondylus. The latter were made from a portion of the sheil just at the junction of its white and rose-colored zones in such a way that one flat face was ailsver rose, while the face of the other side had rose-colored feef but a white upper part. Each of these bifids was drilled with two holes, possibly for suspension on a double cord: the holes enter the white side and emerge through the nmwtop face of the object." (Paulsen 1968:2). -Site: Piriacay

Location: Aniay, Ecuador. Near to the modem town of Paute.

Period: Fornative (Initial PerbdEdy Horizon)

Contents:

A complete necklace consisting of cuentas of Spondylus combined with anthropomorphic and zoomorphic figures of turquoise in the same necklace (Bruhns l989:63).

Comments:

There are no local sources of turquoise in the area or on the coast

Other shells found at Pirincay include: Pinctada mazatlanicu (mother-of- pearl), Strornbus, Conus xintznez and severai others (Bnihns 1989:66).

Based on the fiequency of imported ceramics, Pirincay can be related to the cultures at Cerro Narrio and Chaullabamba, and other extemal contacts. The evidence of long-distance exchange begins in the lower levels with a date of BC 13~1400(Bnihns1989:61).

Excavations in Pirincay indicate that more than 30 bead and crystal workshops were in use on the site during al1 thne periods (Bnihns 1989:70).

Because of the high number of workshops in Pirincay, production would oversupply domestic demand, it is possible that Pirincay was a centre of trade and artisans (Bruhns 1989:72). -Site: Playa ~MiIkr

Location: About 1.5 km south of ArÏca, Chile

Penod: Late Horizon

Contents:

A single red Spondyus bead similar to that of Alacrh Island beads (Bird 1943).

Comrnents: - The site included a midden and graves, where Spondyhs was almost non- existent apart From the single bead.. -Site: Potrem de Payogasta

Location: Caichaqui Valley, northwestem Argentina

Period: Late Horizon/inca

Contents:

+ Sumptuary goods include Spondylus and turquoise beads found in the production debris of Inca elite households (Earle 1994:450).

Comments:

The purpose behind the Inca involvernent in craft industries in the Calchaqui Valley was the production and expropriation of weaith (Earle 1994:443).

Materials such as mica, turquoise and copper were available within the Calchaqui region, but sumptuary goods such as Spondylus beads would not have been carried to northwestem Argentina to be worked. Their presence is more likely due to being carried as personal adomrnents by Inca oficials.

According to Earle (1 994:450) only one recovered bead was made of Spondylus. -Site: Puémape

Location: Cupisnique. Peru, between the Valleys of Jequetepeque and Chicama

-Period: Initial Penod (Early Formative, BC 1000-BC 500)

Contents:

Some of 40 bodies in burials were adomed with collars of Spondylus beads, rnother-of-pari, lapis lanrli, turquoise, and cry stal (Hem 1993 1246).

Comments:

Also a Stone Cupisnique vesse1 with the representation of anthropomorphic eagles with feline characteristics and serpent-like arms holding Spondylus shells while tetuning to the water (Elera 1993:249).

According to Elera (1993:252) the interchange that brought Spondylus and Strombus to Puémape may have been responsible for the ceramic tradition similarities of Cupisnique, Chorrera and Machalilla in Ecuador. -Site: Punkuri

Location: Nepefia Valley, Pem.

Period: Initiai Period

Contents:

The burial of a decapitated fernale included: a kilo of turquoise beads of various forms and sizes, located around the body at waia level; a Sirombus shell trumpet; a pair of Spondjlus princeps shells; 100 examples of Scutalus proteus, and the skeletal remains of guinea pigs and birds that had been fragmentecl and pulverized (Proulx MS:38).

Comrnents:

Punkuri bajo (PV 3 1- 1O) was a terraced artificial mound identified as a temple. The buriai described above was found under the floor of the terrace and indicates a sacrifice placed there as an offenng.

nie dating of the site relies in part on the construction use of conical adobes which were used in the Initial Period and Eariy Horizon.

The site has been reiated to Chavin de Huhtar and Cerro Blanco by Proulx through the presence of Chavin culture artïfacts such as: Stone mortar and pestle; a Strombus trumpet; and feline sculptures in clay.

Another relationship not mentioned by Roulx is the presence of Spondylus shells with a fernale skeleton. A characteristic trait and relationship that is based on the female/Spondylus metaphor found at Chavin. Location: Tumbes River, Pem Approx. 28 km dand hmTumbes.

Period: Inca/Late Horizon

Contents:

Cache of worked Spondylus princeps hinge pieces (personal observation 1993).

Comments:

Possible workshop location and tumbo on a llama caravan route from Cabeza de Vaca through the Amotape Mountains to the central highlands.

See Cabeza de Vaca -Site: Saiaago

Location: Manabi Coast, Ecuador

Period: Edy Formative (various occupations hmthe Preceramic through Late Horizon)

Contents:

Spondylus shell and debris fiom workshops found in levels dating ftom the Machalil la Phase, Chorrera, Guangala and Mantefio Phase (Norton, L unnis and Nailing 1 983).

Comrnents:

Salango appears to have been a centre of Spondylus activity and trade. It was identified by Presley Norton (1986) as a Spondylus workshop, but this should not be seen to diminish the importance of the 17 other levels of cultural occupation existing on the site that are not related to Spondylus activity (Richard Lunnis 1994: personal communication). Location: Santa Elena PenUisu1a, Ecuador

Period: Regional Development (Early Intemediate Period)

Contents:

Spondylus with the coloured margins removed. No immature specimens found (Padsen 1974597)-

Comments: -Site: Sipain

Location: Lambayeque Valley, Peru

Period: Early Intermediate Pd(appmx 200-300 AD)

Contents:

The warrior-priest was entombed wah Spomiylus shells (cornpiete haif-shells) at his feet, and with pectoral coverings of Spondylus shell and copper beads (Alva 1988:s 10-549).

Comments:

During the Early Intermediate Period only Mochica (Moche) burials included Spondylus offenngs such as those fiom Siph (Shirnada 199 1: LI) -Site:

Location: Moche Valley, Peru.

Period: Late Intermediate Period

Contents:

In Cuadro ID, located in the southwest sector of the site, the first level included fragments of painted fnezes, fragments of textiles, and shells including Spondylus and Sirombus. The second level contained worked shell used as inlays for wooden idols. Level 6 contained an id01 decorated with shelt inlays in the shape of fish (Navarro Santander 1986:27-29).

In Cuadro IC worked shells and inlays of mothersfipeari were found.

Another id01 (a person 57 cm long and 20 cm wide made of algarobbo wood, one of four found) recovered has inlays of moth-o6pearl and was adorned with Spondylus of red and white (Navarro Santander 1986:32,4O).

Comments:

According to Navarro Santander (1986:43-44), "a considerable quantity of shell was recovered in disturbed context- ïhe majonty of which have been worked and fonned part of offerïngs. Some of these remains were use as inlays in the idols fond on the site. The species that can be defined are Spondylus and Strornbus. These species have ken worked in various forms, like, pelicans, fish, circles, rectangles and others as inlays. This material, with the exception of the inlays, is not finished or polished. These materials have been used in the manufacture of beads" -

"Estos elernentos (restos malacologicos) ha sido hailado en una cantida considerable; los mismos que se encuentran en un mismo contexto que la rnayoria de elementos hallados durante nuestras excavaciones, en contexto disturbado para ser mis especifices. Presentan, en su mayeria, huellas de haber sido trabiijades, para luge haber fonnado parte de algunas ofrendas. Algunos de estes restos fueron utilizados para las incrustaciones de los idolos que aparecen en el sitio. Las especies que se han podido definir son el Spondillus y otra que parece ser Et Strombus. Estas especies han sido talladas en formas variadas, como pelicanos, peces, circuios, rectanguios y otres, los que han sido utilizados en las incrustaciones. Este material, a excepcion de las incnistiaciones, no presentan un buen acabados, es decù no han sido pulidos. Estos materiales han sido utilizados en la fabrication de cuentas- " -Site:

Location: Valdivia Valley, Manabi Goa* Ecuador

Period: Formative (Precerarnic -Initial Pd)

Valdivia Phase Period C BC 2050 - E3C 1450 Period B BC 2350 - BC 2050 PeriodA EK3550-BC2350

Contents:

Excavation of G-3 1 included Spondylus abraders, poiishers, and beads (Meggers, Evans and Estrada 1965:3 7-3 8).

Comments:

Evidence of cutting by sandstone saw, and also the use of the shells hinge area. Some of the shells were identified by the distinctive coloured lip of the Spondv1usS

0 The recovery of Spondylus shells at G-3 1 was separated into time periods by quantity and percentage of overall sheil specimens (Meggers, Evans and Estrada 1965: 185) as follows:

Period C 235 specimens 15.0 % Period B 97 specimens 16.0 % Period A 8 specimens 1.1 % -Site: VddMi(0GSE-MA-172)

Location: Manabi Coast, Ecuador

Period: Formative -Regionai Developrnent - Integration (various occupations fiom the Preceramic through Late Horizon)

Contents:

Perforated bi-conical cuentas of red or pink Spondylus. Also a vesse1 containhg a quantity of perforated Spondylus beads (Stothert 19932324).

Comments:

The looted remains of some Valdivia sites show evidence of Spondylus princeps and calcifer in rounded, polished, and triangular shapes (personal observation) -Site: V-124

Location: Vini Valley, nonh coast of Pem.

Period: Late Intermediate PeriodLate Horizon (Chimu - ChimdInca)

Contents:

A total of 413 sherds; beads of shell, stone and copper, a few animal bones; and 25 Spondyius shells (Collier 1955:44).

Comments:

Site V-124 is a walled compound that contauied courts, platforrns, rooms and quaclrangles. The contents were recovered from the fil1 near the floors of a U- shaped room and uicluded fragments of Inca vessels. - Sites in the Virb Valley are also connected to the Chicama Valley and Moche Valley during the Chimu-Inca phase of occupation. -Site: V-304

Location: Virit Valley, north Coast of Pem-

Period: Late Intermediate P&od/Late Horizon (Chimu - Chimuha)

Contents:

From Burial 1: A flexed, upright body of indeterminate sex, with a Chimu plain pot and two artïculated Sporidyltcs shells (Collier 1955:47).

Comments:

Site V-304 is a small cemetery located about 500 metres West of V-124 (Collier 1955:47).

Sites in the Vih Valley are also connected to the Chicarna Valley and Moche Valley during the Chirnu-Inca phase of occupation. A~~ndixB

Table 1. Toponyms ContaÏning Variations of the Word 'Mullu' in Ecuador

-- . .- - -- - Nombre Accidente Geo-&CO Provincia Coordenadas Geoesaficas

Mullo Corral Localidad Bolivar LW 78 58 17 LS 013008 Al NE de la Ciudad de Guaranda a 10 kl., ait 3.400 ms-mm-

MuUo Pungu Sitio ba~ LW 79 42 30 LS 03 02 24 Al NE de la Poblacion de Camico Ponce Enrique%a 4 kl-

Cerro Chimborazo LW 78 53 00 LS 0157at Al NE de la Poblacion de Pallatanga a 10.1 kl.,ait 3.856 m.s.n.m.

Loma Chimborazo LW SS 53 36 LS O1 562s Al NE de la Poblacion de Pallatanga a 10.1 kl., ait. 3.856 m.s.n.m.

Muliuleo Loma Cotopaxi LW 78 56 42 CS 01 15 16 Al NE de SUniaîug a3.9kl., alt 2.960 ms-n-m.

Muiiunuma Localidad Loja LW 79 55 16 LS 0407% Al SE de la Ciudad de Celica a5 kl.,alt. 1.540 - - Nombre Accidente Geografico Provincia Coordenadas Geo.&cas

Mullupungu Rio El Oro LW 79 38 07 LS 03 17 30 Afluente del Rio Jubones, ai NO de la Pobiacion de Ushcurrumi-

Quebrada Tunprahua LW 78 30 12 IS O1 15 56 SE une kl ai Rio Cacamaca, ai NE de la Poblacion de Sirniamg. hm: Indice Toponimica de la Republic del Emador Tom0 W (LL-M-N-N-O), compilai by the Instituto Geografico Militar Del Ecuador. Table 2. Spondylus Materiai in the 'Museo de Jacinto Jijon y Caarnaiio' (Quito) From Cern Narrio and surrounding area.

Museum Colour of Sizein mm Description Number outershdl Lx Wx D

1. 2081 Orange 62 24 tapered carved figurine wth a flat head

3. 1881 white 82 22 25 taperd figurine wl head, eyes, haeds

4. 2884 PUV~~ 64 2 1 23 tapered, carved figure with head, body and bands but no ms.One eye is inlaid with translucent material. possibly 'madreperia'.

6. 1886 ~Wle 77 35 15 same as no- 5 but with feet instead of ending in a point

7- 684 white 89 33 not Spondjitlp material, although the item is a carved, tapered figmïne. Looks like the S'n@ius figures with the outer shell moved but has diagond cmed (natllral?) lines instead

7740 Orange 44 10 10 cuenta with driiled hoies on sides.

9. 2014 ota~~&e/red ? 17 18 fragment ofa head w/ han& and eyes

10, 1864 white 70 24 34 figurine with an extra head carved on to main head- The red has been removed hmthe shell. Corresponds to Uhle's Fig. 136.

11. 1887 white 80 27 26 figurine w/ extra heads in the corners of the main head Uhle Fig. 138 on left of group. Museum Colour of Size in mm Description

white fragment ofa tapered tigurine.

ewle flat head figurine-

purpie flat head figurine with anns showing on the side,

puikish trïangular shaped fiagment with extra heads on the forehead Head appean to be carved out of the shell hinge area.

redocange flathead figurine with ears.

white tapered figurine w/ incised lines on back, probably not SpndyIus, meas # 7. the incised lines are 1 mm in width separated by 4 mm ndges.

white

red figurine with ears on head and belt markings at wain

brown figurine w/ arms and han& pointïng upward Udike other figures.

purple fiagment of figurine w/ an extended chin

white figurine w/a flat head, arms and han&-

white figurine possibly holding a child ühle Fig- 135.

white figurine Museum Colour of Size in mm Description Number outer shell Lx Wx D

24, 1997 orange 25 4 3 polished figurine

25, 684 37 5 3 figurine

26, 2056 white 40 6 3 w/feet, cars, and anns. Polished ail over.

47 8 5 outer shell still there, w/ feet, hips, and legs.

28, 604 white 65 20 6 possibly not Spn@ius, similar to #'s 7 & 17.

Table 3. Shells were collected by Jijon y Caamaiio at various sites in Ecuador and Peru ïhere are 48 complete half shells, both princeps and calcifer, with no site reference information,

Museum Colow of Size in mm Description Number outer shell Lx Wx D

29, 4255 120 115 32 halfSpnd'yIus shell

30. 5782 white hdf shell with outer layer removed polished

195 173 51 haifshell largest in the collection

32. 04712 pinkked 84 83 5 half sheli smallest in collection

33. nia pink 101 99 8 half shell w/ outer shell polished off dom to the centre material.

34. nia Worange 15515533 haifshell Table 4. Collection of polished shell in various shapes and sizes with an -tionai 23 polished end and hgment pieces that show evidence of breaks after they were polished

Museum Colour of Size in mm Nurnber outershell Lx Wx D

polished edge piece with a drilied hole in it and an unpolished back of outer shell matehL

broken piece of SpondyIus-

edge piece, trapezoidd shape, w/ white sheIl inside.

fiagrnent of polished outside layer materiai-

40. nia orangeiwbite 27 27 27 9 square Cüentar

41. nia orangdwhite 18 10 12 3 cuentas drilleci with a 5 mm opening that tapers towards the centre.

-pool tapers down to 11 mm W/ a centre hole of 3 mm.

trapezoidai piece tapering to a width of 21 mm, usïng only the outer layer of material, the inside is polished.

trapezoidal piece tapering to 19 mm, the outside is mugh and white, w/ a polished inside of white and red Museum Colour of Size mmrn Description Nwnber outershd Lx Wx D

48. nia orange 86 29 9 poiished on both sides the mentis wedge saaped and stops at the edge where the red margin has bem cut off

Orange 74 31 3 polished inside made fiom the outer material layer, rectangular; shaped with a corner missing-

50. nia OMS 59 22 9 wedge shaped piece of polished outer material

51. da white/orange 27 18 1 e-1, balf whitehiiforange, flared at bottom to 20 mm-Polished dom to I mm in thickness. Colour position indicates that the item was carved fiom a piece taken vertïcally out of the shell.

52. nia omnge/white 30 12 1 eanpool, orange w/ white on the length side indicating ththe material was wdhorizontally .

53. nia omgdwhite 58 21 17 rounded piece of orange Sp0ndyiu.s with white stredcs- Solid, polished and tapers to 16 mm, shaped like a tusk or leg. Table 5. Spondylus Figwïtas From the Banco Central Collection on Display at the Casa De Cultura, Quito, Ecuador (1 2 May 1994).

Museum Colour of Size in mm Description Number outer shell Lx Wx D

1- 173-40-55 white da Figrnine of polished white inside iayer, w/ trnadreperIa' inlays in the eyes. Two anns with one extended across the stomaçh. Very chubby cheeks and forehead (see Figure 3 right).

Figurine w/ red shell backing. Thick and rough loohg wi a chubby face and forehead Also indications of a bel.

3. 474-4 ? white nia Figurine with no red sheil polished on both sides. Has a flat face, fine features, a belt, is quite thin, and has a duck on each side of the forehead (see figure 3 le@. Mayan Use of Spmdyllrcr in Mesoamerica

In Mesoamerica the ritual use of Spondylus was shared by the Maya and Aaecs who used it as a primitive currency with value as a unit of economic exchange. for ritual purposes (Kolb 198723; Richards and bekelman 1937: 167), and as a visible reminder of hard work, status. and power (Norton 1986: 135; Salomon 1986: 124; ScheIe and Freidel 1990:93). Contact with the Pacific Coast appears to have peaked during the Eariy Classic at a time when Teotihuacan trade also appeared in ceremonid contexts (Kolb 1987; Moholy-Nagy 1963:8 1). During this Early Classic perïod and later in the Early Postelassic, it is Spondylus calcifr that is found in association with offerings in areas like Belize and Monte Albh, while duri-ng the Classic period Spondylzus princeps is found in the Maya sites at Amatitlan, San José, Guayatan. Uaxactiin, Coph and Tikal (L. Feldman 1972:130- 13 1: Willey 1972:22 1). At these sites the Spondylus shell was scraped on the interior to bnng out the red colour of the shell, and used as beads, bangles, mosaic elements, figurines of humans and other, cut, carved, incised or painted small worked objets (Moholy-Nagy 1963:70). The difference in species utilization nom calfifer to princeps and back to calcifer may be an indication that princeps was the choice of the elite during the period of their greatest control and power. Once the Classic period began its decline the ability to procure the most valuable species may also have waned. An altemate possibility would be that after the Classic, dedine in use of Spondjdus princeps coincideci w*h the increased use and demand for Spondvlus in the Andes during the Eariy Intemediate Period. Perhaps this is a clue that Spondylus sheil was being sent south to the Andes. The importance of SpondyCus to mesoamerican cultures is found in its near exclusive occurrence in buriais which suggests its symbolic value (Moholy-Nagy 1963 :80y1 989: 142). In Bunal 85 at Tikal, Spondylus was found with stingray spines in the cinnabar irnpregnated bundle of a headless, thighless corpse (Schele and Freidel l9W: 135). Since both Cinnabar and Spondylus are red in colour. and stingray spines were used for bloodletting, they may be related symbolically. Most of the Spondylus examples found at Tikal were probably Spondylus calcifer, but it is interesting to note that marine pearl oyster was also found. including Pinctada mazatlantca and Pteria stema (Moholy-Nagy 198%142). Excavations at Pacbitun, Belize, by the Trent University-Pacbitun Archaeological Project recovered an importeci Spondylus princeps shell dated to the Terminal Classic Period (AD 700-900) (Heaiy 1990:256). Also at Pacbith, Coc phase excavations (AD 550-700) revealed a tomb housing a 45-year old male skeleton sprinkled with red octue with a painted Spondylus shell positioned over the skull (Healy 1990:257). At the site of Altar de Sacrificios Spondylus beads were found in buriais and used in composites of shell and jadeite dong with mother-of-pearl (Willey 1972:223-227). The association of mother-of-pearl with Spondylus in Mesoarnerica is similar to the symbolic relationship between the two shells that appears in the Andes. The connection of mesoamerican cultures to the trade network was based on a demand for elite goods. The ability to obtain foreign goods by trade was critical to the elite Maya As the elite groups collapsed so did the use of etite materials, perhaps in part due to a lack of those materials. Either way, once the elite groups disappeared the area may have been excluded fiom the trade network. a situation that could have prevented recovery from the coilapse (Cowgill 197958; Moholy-Nagy 1989:139). During Classic Maya times, trade routes aiready existed for the transport of obsidian fiom EI Chayal and ktepeque in Guatemala through the Maya Iowlands by river and overland that could also transport goods from the coast (Healy, McKiilop and Walsh l984:416). Trade to the Arne- south-west for supplies of turquoise during the Classic, Late Classic, and Post Classic (Harbottle and Weigand 1992) and the shell exchange networks from the Pacific (Pires-Ferreira 1982) indicate that shorî-distance and long-distance trade was well developed by mesoarnencan cultures, Of the 18 panarnic species of marine shells found in Mayan sites, none of them came from the coasts of Chiapas or Guatemala which are closest to the Mayan area as they were restricted to habitats north of Oaxaca and south of Costa Rica (L. Feldman 1972: 129,130)- This indicates thah while various May an groups in both the lowland and highland areas were using sea shells fiom the Pacific coast, they were involved in obtaining them from Pacific coastal areas through a regional network of trade.